Benchmark Almanac

Open Arabic LLM Leaderboard v1

silma-ai SILMA-9B-Instruct-v1.0 70.6% tops its leaderboard Qwen2.5-72B 69.1% leads the models we track
Not in model cards

Models are scored on AlGhafa with native and translated sets, the ACVA and AceGPT sets, and translated MMLU and EXAMS. The board ranks by the average.

Version 2 (2025-02) removed saturated and machine-translated tasks and fixed an AlGhafa scoring bug. Scores are not comparable across versions.

Released
May 2024
Kind
Knowledge in many languages
Level
School
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Strongest model tested
Qwen2.5-72B #165 on the capability index
Last model added
January 2025
Scores refreshed
November 12th, 2026
734
Entries on its own leaderboard, 55 models we track

silma-ai SILMA-9B-Instruct-v1.0 tops the leaderboard with 70.6%

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Qwen2.5-72B, at 69.1%.

The newest model with a score here came out in January 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

734 entries. Scroll sideways for the rest.

  1. 70.6 silma-ai SILMA-9B-Instruct-v1.0
  2. 69.2 MaziyarPanahi calme-2.1-qwen2.5-72b
  3. 69.1 Sakalti ultiima-72B
  4. 69.1 Replete-AI Replete-LLM-V2.5-Qwen-72b
  5. 69.1 MaziyarPanahi calme-2.2-qwen2.5-72b
  6. 69.1 rombodawg Rombos-LLM-V2.5-Qwen-72b
  7. 69.1 Qwen2.5-72B
  8. 68.1 zetasepic Qwen2.5-72B-Instruct-abliterated
  9. 67.1 paloalma TW3-JRGL-v2
  10. 67.0 davidkim205 Rhea-72b-v0.5
  11. 66.5 AXCXEPT EZO-Qwen2.5-72B-Instruct
  12. 65.9 freewheelin free-evo-qwen72b-v0.8-re
  13. 65.8 MaziyarPanahi calme-2.2-qwen2-72b
  14. 65.1 dnhkng RYS-XLarge
  15. 65.1 arcee-ai Arcee-Nova
  16. 65.0 dnhkng RYS-XLarge-base
  17. 65.0 MTSAIR MultiVerse_70B
  18. 64.7 anthracite-org magnum-v1-72b
  19. 64.7 Undi95 MG-FinalMix-72B
  20. 64.7 MaziyarPanahi calme-2.1-qwen2-72b
  21. 64.7 dfurman Qwen2-72B-Orpo-v0.1
  22. 64.7 MaziyarPanahi Qwen2-72B-Instruct-v0.1
  23. 64.7 alpindale magnum-72b-v1
  24. 64.6 abacusai Dracarys-72B-Instruct
  25. 64.4 silma-ai SILMA-9B-Instruct-v0.1
  26. 64.4 Nexusflow Athene-70B
  27. 64.4 mssqpi G3
  28. 64.3 silma-ai SILMA-9B-Instruct-v0.1.0
  29. 64.3 silma-ai SILMA-9B-Instruct-v0.1.1
  30. 64.0 tanliboy lambda-qwen2.5-32b-dpo-test
  31. 64.0 Sakalti ultiima-125B
  32. 63.7 anthracite-org magnum-v2-72b
  33. 63.6 Qwen2-72B
  34. 63.5 paloalma Le_Triomphant-ECE-TW3
  35. 63.4 abacusai Smaug-Qwen2-72B-Instruct
  36. 63.1 pankajmathur orca_mini_v7_72b
  37. 63.1 Qwen/Qwen2.5-32B-Instruct
  38. 63.0 paulml ECE-ILAB-Q1
  39. 63.0 Sakalti ultiima-32B
  40. 62.9 Replete-AI Replete-LLM-V2.5-Qwen-32b
  41. 62.8 VAGOsolutions Llama-3-SauerkrautLM-70b-Instruct
  42. 62.7 Sakalti ultiima-78B
  43. 62.6 dfurman CalmeRys-78B-Orpo-v0.1
  44. 62.4 sometimesanotion Qwen2.5-14B-Vimarckoso-v3
  45. 62.3 sometimesanotion Lamarck-14B-v0.6
  46. 62.3 Sakalti ultiima-78B-V2
  47. 62.3 hotmailuser QwenSlerp2-14B
  48. 62.2 AXCXEPT EZO-Qwen2.5-32B-Instruct
  49. 62.1 Sakalti SakaMoe-3x14B-Instruct
  50. 62.0 tanliboy lambda-qwen2.5-14b-dpo-test
  51. 61.8 huihui-ai Qwen2.5-32B-Instruct-abliterated
  52. 61.7 sometimesanotion Lamarck-14B-v0.7
  53. 61.7 Slim205 Barka-9b-it-v02
  54. 61.7 airev-ai Amal-70b-v2
  55. 61.6 airev-ai Amal-70b-v1.1
  56. 61.6 migtissera Tess-v2.5.2-Qwen2-72B
  57. 61.5 airev-ai Amal-75b-v3
  58. 61.5 byroneverson gemma-2-27b-it-abliterated
  59. 61.4 MaziyarPanahi calme-2.4-llama3-70b
  60. 61.3 Lambent qwen2.5-reinstruct-alternate-lumen-14B
  61. 61.3 v000000 Qwen2.5-Lumen-14B
  62. 61.2 MaziyarPanahi calme-2.2-llama3-70b
  63. 61.2 rombodawg Rombos-LLM-V2.6-Qwen-14b
  64. 61.1 Replete-AI Replete-LLM-V2.5-Qwen-14b
  65. 61.1 Sakalti ultiima-14B
  66. 61.1 v000000 Qwen2.5-14B-Gutenberg-1e-Delta
  67. 61.1 airev-ai qwen-14b-v1
  68. 61.1 Qwen/Qwen2.5-14B-Instruct
  69. 61.0 v000000 Qwen2.5-14B-Gutenberg-Instruct-Slerpeno
  70. 60.9 VAGOsolutions Llama-3.1-SauerkrautLM-70b-Instruct
  71. 60.9 FiditeNemini Unhinged-Qwen2-72B
  72. 60.9 yellowtown 7B-v0.2
  73. 60.8 MaziyarPanahi calme-2.3-llama3-70b
  74. 60.8 NousResearch Hermes-3-Llama-3.1-70B
  75. 60.6 google/gemma-2-27b-it
  76. 60.2 elmrc JuhainaG2
  77. 60.1 MaziyarPanahi Llama-3-70B-Instruct-DPO-v0.2
  78. 60.0 mmnga Llama-3-70B-japanese-suzume-vector-v0.1
  79. 59.9 airev-ai Jais-Inception-70b-V1.2.2
  80. 59.9 meta-llama/Meta-Llama-3-70B-Instruct
  81. 59.8 failspy llama-3-70B-Instruct-abliterated
  82. 59.5 dnhkng RYS-Gemma-2-27b-it
  83. 59.5 airev-ai qwen-14b-v2
  84. 59.4 airev-ai Jais-Inception-70b-V1.1
  85. 59.3 recoilme recoilme-gemma-2-9B-v0.4
  86. 59.1 cognitivecomputations dolphin-2.9.2-qwen2-72b
  87. 59.0 migtissera Tess-v2.5-Qwen2-72B
  88. 58.8 Qwen2.5-14B-Instruct-1M
  89. 58.8 abacusai Smaug-72B-v0.1
  90. 58.6 elmrc juhainaG
  91. 58.5 princeton-nlp gemma-2-9b-it-SimPO
  92. 58.5 elmresearchcenter juhaina
  93. 58.4 gbueno86 Meta-LLama-3-Cat-Smaug-LLama-70b
  94. 58.2 Nohobby Q2.5-Qwetiapin-32B
  95. 58.2 anthracite-org magnum-v3-27b-kto
  96. 58.0 nvidia Llama-3.1-Nemotron-70B-Instruct-HF
  97. 58.0 Daemontatox Mawared_GS
  98. 58.0 airev-ai Amal-70b-v5
  99. 57.9 arcee-ai Virtuoso-Small
  100. 57.8 maldv Qwenstein2.5-32B-Instruct
  101. 57.8 abhishek autotrain-llama3-70b-orpo-v1
  102. 57.6 gbueno86 Meta-LLama-3-Cat-A-LLama-70b
  103. 57.5 abhishek autotrain-llama3-70b-orpo-v2
  104. 57.4 VAGOsolutions SauerkrautLM-Nemo-12b-Instruct
  105. 57.2 UCLA-AGI Gemma-2-9B-It-SPPO-Iter3
  106. 57.1 MaziyarPanahi calme-2.3-llama3.1-70b
  107. 57.0 UCLA-AGI Gemma-2-9B-It-SPPO-Iter1
  108. 57.0 airev-ai emirati-14b-v2
  109. 57.0 google/gemma-2-9b-it
  110. 57.0 Ashmal MBZUAI-ORYX-new
  111. 57.0 airev-ai Jais-Inception-70b-V1.2
  112. 57.0 recoilme Gemma-2-Ataraxy-Gemmasutra-9B-slerp
  113. 57.0 UCLA-AGI Gemma-2-9B-It-SPPO-Iter2
  114. 56.9 MaziyarPanahi calme-2.2-llama3.1-70b
  115. 56.9 cognitivecomputations dolphin-2.9.1-llama-3-70b
  116. 56.9 rombodawg Rombos-LLM-V2.6-Nemotron-70b
  117. 56.8 inceptionai jais-adapted-70b-chat
  118. 56.8 elmrc JuhainaG9B
  119. 56.8 failspy Llama-3-70B-Instruct-abliterated-v3
  120. 56.8 shenzhi-wang Llama3-70B-Chinese-Chat
  121. 56.8 mssqpi Waie1
  122. 56.8 TareksLab Thalassic-Alpha-LLaMa-70B
  123. 56.8 airev-ai Amal-70b-v4
  124. 56.6 hotmailuser Deeepseek-QwenSlerp4-32B
  125. 56.6 sambanovasystems SambaLingo-Arabic-Chat-70B
  126. 56.5 recoilme recoilme-gemma-2-9B-v0.2
  127. 56.4 Steelskull L3-MS-Astoria-70b
  128. 56.4 MaziyarPanahi calme-2.3-qwen2-72b
  129. 56.4 CohereForAI/c4ai-command-r-plus
  130. 56.2 meta-llama/Meta-Llama-3.1-70B-Instruct
  131. 56.2 tenyx Llama3-TenyxChat-70B
  132. 55.9 BAAI Gemma2-9B-IT-Simpo-Infinity-Preference
  133. 55.9 inceptionai jais-family-30b-8k-chat
  134. 55.8 Ashmal MBZUAI-oryx
  135. 55.7 migtissera Llama-3-70B-Synthia-v3.5
  136. 55.7 CohereForAI/c4ai-command-r-plus-08-2024
  137. 55.7 princeton-nlp gemma-2-9b-it-DPO
  138. 55.5 recoilme recoilme-gemma-2-9B-v0.3
  139. 55.3 Triangle104 DeepSky-T100
  140. 55.2 airev-ai emirati-14b-v3
  141. 55.2 Gemma 2 27B
  142. 55.0 huihui-ai Qwen2.5-7B-Instruct-abliterated-v2
  143. 55.0 Delta-Vector Odin-9B
  144. 55.0 FreedomIntelligence AceGPT-v2-70B
  145. 54.9 prithivMLmods Qwen2.5-32B-DeepSeek-R1-Instruct
  146. 54.9 Joseph717171 Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
  147. 54.8 Ba2han QwQenSeek-coder
  148. 54.6 FreedomIntelligence AceGPT-v2-32B-Chat
  149. 54.5 core42 jais-30b-chat-v3
  150. 54.5 Orion-zhen Qwen2.5-7B-Instruct-Uncensored
  151. 54.5 anthracite-org magnum-v4-9b
  152. 54.4 Command R (03-2024)
  153. 54.3 Isaak-Carter Josiefied-Qwen2.5-7B-Instruct-abliterated
  154. 54.3 BAAI Infinity-Instruct-7M-Gen-Llama3_1-70B
  155. 54.3 Qwen/Qwen2.5-7B-Instruct
  156. 54.2 Qwen2.5-32B
  157. 54.2 ai21labs AI21-Jamba-1.5-Mini
  158. 54.2 anthracite-org magnum-v3-9b-customgemma2
  159. 54.1 Replete-AI Replete-LLM-V2.5-Qwen-7b
  160. 54.1 TheDrummer Gemmasutra-9B-v1
  161. 54.1 migtissera Tess-2.0-Llama-3-70B
  162. 54.0 jondurbin airoboros-dpo-70b-3.3
  163. 54.0 failspy Meta-Llama-3-70B-Instruct-abliterated-v3.5
  164. 53.9 migtissera Tess-2.0-Llama-3-70B-v0.2
  165. 53.8 jondurbin airoboros-70b-3.3
  166. 53.8 Sakalti Saka-7.6B
  167. 53.8 RDson WomboCombo-R1-Coder-14B-Preview
  168. 53.7 c4ai-command-r-08-2024
  169. 53.7 mistralai/Codestral-22B-v0.1
  170. 53.7 airev-ai Jais-Inception-70b
  171. 53.6 DeepSeek-R1-Distill-Llama-70B
  172. 53.6 arcee-ai Arcee-Spark
  173. 53.5 Daemontatox Llama3.3-70B-CogniLink
  174. 53.5 huihui-ai Qwen2.5-7B-Instruct-abliterated
  175. 53.5 mssqpi ElmGemma2
  176. 53.4 Isaak-Carter Josiefied-Qwen2.5-7B-Instruct-abliterated-v2
  177. 53.4 Qwen2.5-14B
  178. 53.3 Cran-May T.E-8.1
  179. 53.3 dfurman Llama-3-70B-Orpo-v0.1
  180. 53.3 Aya Expanse 8B
  181. 53.1 migtissera Tess-3-Llama-3.1-70B
  182. 53.1 Qwen2.5-Math-72B-Instruct
  183. 53.1 paloalma ECE-TW3-JRGL-V2
  184. 53.1 dnhkng RYS-Llama3.1-Large
  185. 53.0 vicgalle CarbonBeagle-11B
  186. 52.9 inceptionai jais-family-13b-chat
  187. 52.9 Qwen/Qwen2-MoE-57B-A14B-Instruct
  188. 52.8 SicariusSicariiStuff Qwen2.5-14B_Uncencored
  189. 52.7 SeaLLMs SeaLLM-7B-v2.5
  190. 52.7 meta-llama/Meta-Llama-3-70B
  191. 52.6 nitky RoguePlanet-DeepSeek-R1-Qwen-32B
  192. 52.6 meta-llama/Meta-Llama-3.1-70B
  193. 52.5 anthracite-org magnum-v3-9b-chatml
  194. 52.5 Llama 3.1-70B
  195. 52.4 abacusai Smaug-Llama-3-70B-Instruct-32K
  196. 52.4 FreedomIntelligence AceGPT-v2-70B-Chat
  197. 52.4 informatiker Qwen2-7B-Instruct-abliterated
  198. 52.3 CohereForAI/aya-23-35B
  199. 52.3 TeetouchQQ model_mergev2
  200. 52.2 abacusai Smaug-Llama-3-70B-Instruct
  201. 52.2 natong19 Qwen2-7B-Instruct-abliterated
  202. 52.0 xiangxinai Xiangxin-2XL-Chat-1048k-Chinese-Llama3-70B
  203. 52.0 OpenBuddy openbuddy-llama3-70b-v21.2-32k
  204. 51.8 SicariusSicariiStuff Qwen2.5-14B_Uncensored_Instruct
  205. 51.8 failspy Smaug-Llama-3-70B-Instruct-abliterated-v3
  206. 51.7 vicgalle ConfigurableBeagle-11B
  207. 51.7 Qwen2-7B-Instruct
  208. 51.7 DeepMount00 Llama-3.1-8b-Ita
  209. 51.7 arcee-ai Llama-3.1-SuperNova-Lite
  210. 51.6 nyunai nyun-c2-llama3-56B
  211. 51.5 princeton-nlp Llama-3-Base-8B-SFT-SimPO
  212. 51.4 paloalma ECE-TW3-JRGL-V1
  213. 51.4 VAGOsolutions Llama-3.1-SauerkrautLM-8b-Instruct
  214. 51.4 Qwen/Qwen2.5-Math-72B
  215. 51.3 vicgalle CarbonBeagle-11B-truthy
  216. 51.3 NaniDAO deepseek-r1-qwen-2.5-32B-ablated
  217. 51.2 utter-project EuroLLM-9B-Instruct
  218. 51.2 elmrc ElmJuhaina70B
  219. 51.1 core42 jais-30b-chat-v1
  220. 51.0 nbeerbower DeepSeek-R1-Qwen-lorablated-32B
  221. 50.9 princeton-nlp Llama-3-Base-8B-SFT-IPO
  222. 50.8 FreedomIntelligence AceGPT-v2-32B
  223. 50.8 nicoboss DeepSeek-R1-Distill-Qwen-32B-Uncensored
  224. 50.8 shadowml Mixolar-4x7b
  225. 50.8 asas-ai AceGPT-v2-32B
  226. 50.8 axolotl-ai-co romulus-mistral-nemo-12b-simpo
  227. 50.6 Qwen1.5-110B
  228. 50.5 invalid-coder Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp
  229. 50.4 jeonsworld CarbonVillain-en-10.7B-v4
  230. 50.4 kekmodel StopCarbon-10.7B-v5
  231. 50.4 Gemma 2 9B
  232. 50.3 nbeerbower mistral-nemo-wissenschaft-12B
  233. 50.3 FreedomIntelligence AceGPT-v1.5-13B-Chat
  234. 50.2 microsoft/Phi-3.5-mini-instruct
  235. 50.2 FreedomIntelligence AceGPT-v2-8B-Chat
  236. 50.1 Sakalti Saka-3.8B
  237. 50.0 openchat openchat-3.5-0106-gemma
  238. 50.0 prithivMLmods Qwen2.5-14B-DeepSeek-R1-1M
  239. 50.0 upstage SOLAR-10.7B-Instruct-v1.0
  240. 50.0 vicgalle ConfigurableSOLAR-10.7B
  241. 49.9 migtissera Tess-v2.5-Gemma-2-27B-alpha
  242. 49.9 RDson CoderO1-DeepSeekR1-Coder-32B-Preview
  243. 49.9 ai21labs Jamba-v0.1
  244. 49.9 anthracite-org magnum-v2.5-12b-kto
  245. 49.9 Sakalti Saka-24B
  246. 49.8 NLPark Shi-Cis-Kestrel-uncensored
  247. 49.8 vicgalle Configurable-Hermes-2-Pro-Llama-3-8B
  248. 49.8 llm-mle ara_mle_pretrained_8b
  249. 49.8 llm-mle arb_mle_pretrained_8b
  250. 49.7 arcee-ai Llama-Spark
  251. 49.7 princeton-nlp Llama-3-Base-8B-SFT-RDPO
  252. 49.7 MarinaraSpaghetti NemoReRemix-12B
  253. 49.7 Qwen/Qwen1.5-110B-Chat
  254. 49.7 AALF gemma-2-27b-it-SimPO-37K-100steps
  255. 49.6 princeton-nlp Llama-3-Base-8B-SFT-DPO
  256. 49.4 Magpie-Align Llama-3-8B-Magpie-Align-v0.3
  257. 49.3 grimjim Llama-3-Instruct-8B-SPPO-Iter3-SimPO-merge
  258. 49.3 MaziyarPanahi calme-2.3-qwen2-7b
  259. 49.2 AALF gemma-2-27b-it-SimPO-37K
  260. 49.2 vilm Quyen-Pro-Max-v0.1
  261. 49.1 Dampfinchen Llama-3.1-8B-Ultra-Instruct
  262. 49.1 openchat openchat-3.6-8b-20240522
  263. 49.1 Daemontatox MawaredT1
  264. 49.0 core42 jais-13b-chat
  265. 49.0 saltlux luxia-21.4b-alignment-v1.2
  266. 49.0 RLHFlow LLaMA3-iterative-DPO-final
  267. 49.0 turboderp Cat-Llama-3-70B-instruct
  268. 49.0 inceptionai jais-adapted-70b
  269. 49.0 Eurdem Defne-llama3.1-8B
  270. 48.9 NousResearch Nous-Hermes-2-Mixtral-8x7B-DPO
  271. 48.9 cognitivecomputations WestLake-7B-v2-laser
  272. 48.9 natong19 Mistral-Nemo-Instruct-2407-abliterated
  273. 48.8 elmrc juhaina
  274. 48.8 elmrc Juhaina28B
  275. 48.8 lightblue suzume-llama-3-8B-multilingual-orpo-borda-top75
  276. 48.7 huihui-ai DeepSeek-R1-Distill-Qwen-32B-abliterated
  277. 48.7 VAGOsolutions SauerkrautLM-Phi-3-medium
  278. 48.7 VAGOsolutions SauerkrautLM-Qwen-32b
  279. 48.7 princeton-nlp Llama-3-Instruct-8B-SimPO-v0.2
  280. 48.6 Qwen2.5-7B
  281. 48.6 lemon-mint gemma-ko-7b-instruct-v0.62
  282. 48.6 princeton-nlp Llama-3-Base-8B-SFT-KTO
  283. 48.5 Danielbrdz Barcenas-14b-Phi-3-medium-ORPO
  284. 48.5 elmrc Juhaina29B
  285. 48.5 Mihaiii Llama-3-pruned-45B-Drobeta-Turnu-Severin
  286. 48.5 princeton-nlp Llama-3-8B-ProLong-64k-Instruct
  287. 48.5 v000000 L3.1-Niitorm-8B-DPO-t0.0001
  288. 48.4 MaziyarPanahi Llama-3-8B-Instruct-v0.10
  289. 48.3 princeton-nlp Llama-3-8B-ProLong-512k-Instruct
  290. 48.3 TIGER-Lab MAmmoTH2-8x7B-Plus
  291. 48.3 MaziyarPanahi calme-2.5-qwen2-7b
  292. 48.3 Qwen1.5-72B
  293. 48.2 jondurbin bagel-8b-v1.0
  294. 48.2 GritLM GritLM-8x7B-KTO
  295. 48.2 inceptionai jais-family-6p7b-chat
  296. 48.1 CohereForAI/aya-23-8B
  297. 48.0 mlabonne NeuralDaredevil-8B-abliterated
  298. 48.0 MaziyarPanahi Calme-4x7B-MoE-v0.1
  299. 48.0 meta-llama/Meta-Llama-3.1-8B-Instruct
  300. 48.0 DeepSeek-R1-Distill-Qwen-14B
  301. 48.0 OpenBuddy openbuddy-deepseek-67b-v18.1-4k
  302. 48.0 HeshamHaroon Testv1
  303. 47.9 MaziyarPanahi Calme-4x7B-MoE-v0.2
  304. 47.9 zhengr MixTAO-7Bx2-MoE-v8.1
  305. 47.9 ibivibiv multimaster-7b-v6
  306. 47.9 inceptionai jais-family-2p7b-chat
  307. 47.8 Kukedlc NeuralLLaMa-3-8b-DT-v0.1
  308. 47.8 princeton-nlp Llama-3-Instruct-8B-RDPO
  309. 47.8 OpenBuddy openbuddy-qwen2.5llamaify-7b-v23.1-200k
  310. 47.8 airev-ai Jais-Inception-7b-V0.1
  311. 47.8 jsfs11 MixtureofMerges-MoE-4x7b-v5
  312. 47.8 kevin009 llamaRAGdrama
  313. 47.8 Kukedlc NeuralSynthesis-7B-v0.1
  314. 47.7 vicgalle Configurable-Hermes-3-Llama-3.1-8B
  315. 47.7 AIDC-AI Marco-LLM-AR
  316. 47.7 princeton-nlp Llama-3-Instruct-8B-SimPO
  317. 47.7 Kukedlc NeuralSynthesis-7B-v0.3
  318. 47.7 saltlux luxia-21.4b-alignment-v1.0
  319. 47.6 MaziyarPanahi Llama-3-8B-Instruct-v0.9
  320. 47.6 nbeerbower gemma2-gutenberg-27B
  321. 47.5 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter3
  322. 47.5 airev-ai Jais-Inception-7b-V0.2
  323. 47.5 allknowingroger MultiverseEx26-7B-slerp
  324. 47.5 Qwen/Qwen1.5-32B-Chat
  325. 47.5 liminerity M7-7b
  326. 47.5 princeton-nlp Llama-3-Base-8B-SFT-SLiC-HF
  327. 47.4 grimjim Llama-3.1-8B-Instruct-abliterated_via_adapter
  328. 47.4 inceptionai jais-adapted-13b-chat
  329. 47.4 elmrc Juhaina-Instruct-v1
  330. 47.4 utter-project EuroLLM-9B
  331. 47.4 jsfs11 MixtureofMerges-MoE-4x7b-v4
  332. 47.3 RubielLabarta LogoS-7Bx2-MoE-13B-v0.2
  333. 47.3 inceptionai jais-family-30b-16k-chat
  334. 47.3 ClusterlabAi GemmAr-7B-v1
  335. 47.3 princeton-nlp Llama-3-Base-8B-SFT-ORPO
  336. 47.3 NousResearch Hermes-3-Llama-3.1-8B
  337. 47.3 EpistemeAI Fireball-Alpaca-Llama-3.1-8B-Philos-DPO-200
  338. 47.2 meraGPT mera-mix-4x7B
  339. 47.2 mlabonne Daredevil-8B-abliterated
  340. 47.2 OpenBuddy openbuddy-qwen2.5llamaify-14b-v23.1-200k
  341. 47.2 sambanovasystems SambaLingo-Arabic-Base-70B
  342. 47.1 dzakwan dzakwan-MoE-4x7b-Beta
  343. 47.1 cognitivecomputations laserxtral
  344. 47.1 Orenguteng Llama-3.1-8B-Lexi-Uncensored
  345. 47.1 TIGER-Lab MAmmoTH2-8B-Plus
  346. 47.1 Qwen1.5-14B-Chat
  347. 47.1 Kquant03 CognitiveFusion2-4x7B-BF16
  348. 47.1 MaziyarPanahi Topxtral-4x7B-v0.1
  349. 47.1 vicgalle Configurable-Llama-3.1-8B-Instruct
  350. 47.1 thariqkhalid ara_mle_sft_300_8b
  351. 47.1 princeton-nlp Llama-3-Instruct-8B-RDPO-v0.2
  352. 47.0 migtissera Tess-3-Mistral-Nemo-12B
  353. 47.0 Ali-C137 Ge1H10M-0000
  354. 47.0 FreedomIntelligence AceGPT-13B-chat
  355. 47.0 princeton-nlp Llama-3-Instruct-8B-DPO-v0.2
  356. 47.0 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter2
  357. 47.0 Slim205 Barka-2b-it
  358. 47.0 EpistemeAI Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R2
  359. 46.9 princeton-nlp Llama-3-Instruct-8B-DPO
  360. 46.9 bunnycore HyperLlama-3.1-8B
  361. 46.9 Kukedlc NeuralExperiment-7b-MagicCoder-v7.5
  362. 46.9 Orenguteng Llama-3.1-8B-Lexi-Uncensored-V2
  363. 46.9 EpistemeAI Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R1-KTO-beta
  364. 46.9 MTSAIR multi_verse_model
  365. 46.9 Kukedlc NeuralSynthesis-7b-v0.4-slerp
  366. 46.9 lightblue suzume-llama-3-8B-multilingual-orpo-borda-full
  367. 46.9 NousResearch Nous-Hermes-2-SOLAR-10.7B
  368. 46.8 automerger YamshadowExperiment28-7B
  369. 46.8 princeton-nlp Llama-3-Instruct-8B-RRHF-v0.2
  370. 46.8 EpistemeAI2 Fireball-Alpaca-Llama-3.1-8B-Instruct-KTO-beta
  371. 46.8 NousResearch Hermes-2-Theta-Llama-3-8B
  372. 46.8 VAGOsolutions SauerkrautLM-Gemma-7b
  373. 46.8 princeton-nlp Llama-3-Instruct-8B-IPO
  374. 46.8 Kukedlc NeuralLLaMa-3-8b-ORPO-v0.3
  375. 46.7 EpistemeAI Fireball-Alpaca-Llama3.1.07-8B-Philos-Math-KTO-beta
  376. 46.7 VAGOsolutions Llama-3-SauerkrautLM-8b-Instruct
  377. 46.7 yunconglong Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B
  378. 46.7 allknowingroger Ph3della5-14B
  379. 46.7 mlabonne Beyonder-4x7B-v3
  380. 46.7 princeton-nlp Llama-3-Instruct-8B-SLiC-HF-v0.2
  381. 46.7 v000000 SwallowMaid-8B-L3-SPPO-abliterated
  382. 46.7 princeton-nlp Llama-3-Instruct-8B-IPO-v0.2
  383. 46.7 DeepMount00 Llama-3-8b-Ita
  384. 46.7 princeton-nlp Llama-3-Instruct-8B-ORPO-v0.2
  385. 46.6 jpacifico Chocolatine-14B-Instruct-DPO-v1.2
  386. 46.6 mlabonne ChimeraLlama-3-8B-v3
  387. 46.6 Weyaxi Einstein-v7-Qwen2-7B
  388. 46.6 MaziyarPanahi calme-2.4-qwen2-7b
  389. 46.6 MaziyarPanahi calme-2.6-qwen2-7b
  390. 46.5 Qwen/Qwen2.5-Coder-7B-Instruct
  391. 46.5 OpenBuddy openbuddy-qwen1.5-32b-v21.1-32k
  392. 46.5 Qwen/Qwen2-MoE-57B-A14B
  393. 46.5 yunconglong MoE_13B_DPO
  394. 46.5 cognitivecomputations dolphin-2.9.2-qwen2-7b
  395. 46.5 MaziyarPanahi calme-2.2-qwen2-7b
  396. 46.5 sophosympatheia Midnight-Miqu-70B-v1.5
  397. 46.5 MaziyarPanahi calme-2.1-qwen2-7b
  398. 46.5 inceptionai jais-adapted-7b-chat
  399. 46.5 Gemma 1.1 7B IT
  400. 46.4 elmrc JuhainaL2
  401. 46.4 uygarkurt llama-3-merged-linear
  402. 46.4 shadowml BeagSake-7B
  403. 46.4 jpacifico Chocolatine-14B-Instruct-4k-DPO
  404. 46.4 mlabonne AlphaMonarch-7B
  405. 46.3 OpenBuddy openbuddy-qwen1.5-32b-v21.2-32k
  406. 46.3 Qwen/Qwen2.5-3B-Instruct
  407. 46.3 princeton-nlp Llama-3-Instruct-8B-KTO-v0.2
  408. 46.3 elmrc JuhainaL
  409. 46.3 MaziyarPanahi Llama-3-8B-Instruct-v0.8
  410. 46.3 phi-3-medium 14B
  411. 46.2 cstr llama3-8b-spaetzle-v37
  412. 46.2 BAAI Infinity-Instruct-3M-0625-Llama3-8B
  413. 46.1 mlabonne NeuralMonarch-7B
  414. 46.1 yunconglong DARE_TIES_13B
  415. 46.1 MaziyarPanahi calme-2.7-qwen2-7b
  416. 46.1 CultriX NeuralMona_MoE-4x7B
  417. 46.1 grimjim Llama-3-Instruct-8B-SimPO-SPPO-Iter3-merge
  418. 46.1 SeaLLMs SeaLLMs-v3-7B-Chat
  419. 46.1 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter1
  420. 46.1 Qwen2.5-Coder-14B-Instruct
  421. 46.0 dnhkng RYS-Llama-3.1-8B-Instruct
  422. 46.0 Gemma 7B
  423. 46.0 inceptionai jais-family-30b-16k
  424. 46.0 Qwen2.5-Coder-14B
  425. 45.9 macadeliccc Samantha-Qwen-2-7B
  426. 45.9 freewheelin free-solar-evo-v0.13
  427. 45.9 leafspark Llama-3.1-8B-MultiReflection-Instruct
  428. 45.8 mlabonne ChimeraLlama-3-8B-v2
  429. 45.8 Magpie-Align MagpieLM-8B-Chat-v0.1
  430. 45.8 maywell Qwen2-7B-Multilingual-RP
  431. 45.8 princeton-nlp Llama-3-Instruct-8B-KTO
  432. 45.8 inceptionai jais-family-30b-8k
  433. 45.8 princeton-nlp Llama-3-Instruct-8B-CPO-v0.2
  434. 45.8 BAAI Infinity-Instruct-7M-Gen-Llama3_1-8B
  435. 45.8 wenbopan Faro-Yi-9B-DPO
  436. 45.8 freewheelin free-solar-evo-v0.11
  437. 45.8 Qwen1.5-32B
  438. 45.7 Phi-3-medium-128k-instruct
  439. 45.7 lightblue suzume-llama-3-8B-multilingual-orpo-borda-top25
  440. 45.7 EpistemeAI Fireball-Meta-Llama-3.1-8B-Instruct-Agent-0.003-128K-code
  441. 45.6 gaverfraxz Meta-Llama-3.1-8B-Instruct-HalfAbliterated-TIES
  442. 45.6 Weyaxi Bagel-Hermes-34B-Slerp
  443. 45.6 Kukedlc NeuralLLaMa-3-8b-ORPO-v0.4
  444. 45.6 HeshamHaroon llama-3-instruct-slerp-arabic
  445. 45.5 freewheelin free-solar-evo-v0.1
  446. 45.5 Ali-C137 Q2AW1M-0000
  447. 45.5 Ali-C137 Q2AW1M-1000
  448. 45.5 Ali-C137 Q2AW1M-1010
  449. 45.5 jondurbin bagel-dpo-34b-v0.2
  450. 45.5 Ali-C137 PG7BB
  451. 45.5 Ali-C137 Q2H10M-0000
  452. 45.5 abacusai Smaug-Mixtral-v0.1
  453. 45.5 Danielbrdz Barcenas-Llama3-8b-ORPO
  454. 45.5 Ali-C137 Q2AW1M-0010
  455. 45.5 Ali-C137 Q2AW1M-1100
  456. 45.4 abacusai bigstral-12b-32k
  457. 45.4 chujiezheng Mistral7B-PairRM-SPPO-ExPO
  458. 45.4 cloudyu TomGrc_FusionNet_34Bx2_MoE_v0.1_DPO_f16
  459. 45.4 mssqpi ALQLora2
  460. 45.4 Ali-C137 Q2AW1M-0100
  461. 45.3 ibivibiv llama-3-nectar-dpo-8B
  462. 45.3 2A2I-R L500MT
  463. 45.3 2A2I-R L100MT
  464. 45.3 cloudyu Mixtral_34Bx2_MoE_60B
  465. 45.3 meta-llama/Meta-Llama-3-8B-Instruct
  466. 45.3 Magpie-Align Llama-3-8B-Magpie-Align-SFT-v0.3
  467. 45.3 2A2I-R L200MT
  468. 45.2 2A2I-R L1000MT
  469. 45.2 Magpie-Align Llama-3.1-8B-Magpie-Align-SFT-v0.1
  470. 45.2 core42 jais-30b-v3
  471. 45.2 GritLM GritLM-8x7B
  472. 45.1 cloudyu Yi-34Bx2-MOE-200K
  473. 45.1 UCLA-AGI Mistral7B-PairRM-SPPO
  474. 45.1 UCLA-AGI Mistral7B-PairRM-SPPO-Iter2
  475. 45.1 Magpie-Align MagpieLM-8B-SFT-v0.1
  476. 45.0 princeton-nlp Llama-3-Base-8B-SFT-CPO
  477. 45.0 BAAI Infinity-Instruct-3M-0625-Qwen2-7B
  478. 45.0 abacusai Slerp-CM-mist-dpo
  479. 44.9 princeton-nlp Mistral-7B-Instruct-RRHF
  480. 44.9 failspy Phi-3-medium-4k-instruct-abliterated-v3
  481. 44.9 cloudyu Yi-34Bx2-MoE-60B
  482. 44.9 UCLA-AGI Mistral7B-PairRM-SPPO-Iter3
  483. 44.8 UCLA-AGI Mistral7B-PairRM-SPPO-Iter1
  484. 44.8 princeton-nlp Llama-3-Base-8B-SFT-RRHF
  485. 44.8 nvidia Llama3-ChatQA-1.5-8B
  486. 44.8 princeton-nlp Llama-3-Instruct-8B-RRHF
  487. 44.7 princeton-nlp Mistral-7B-Instruct-SimPO
  488. 44.7 princeton-nlp Llama-3-Instruct-8B-CPO
  489. 44.6 failspy Codestral-22B-v0.1-abliterated-v3
  490. 44.6 silma-ai SILMA-Kashif-2B-Instruct-v1.0
  491. 44.6 cloudyu Yi-34Bx2-MoE-60B-DPO
  492. 44.6 vicgalle Configurable-Llama-3-8B-v0.3
  493. 44.5 princeton-nlp Mistral-7B-Instruct-DPO
  494. 44.5 cognitivecomputations dolphin-2.9.3-llama-3-8b
  495. 44.5 cognitivecomputations dolphin-2.9.4-llama3.1-8b
  496. 44.4 cognitivecomputations Llama-3-8B-Instruct-abliterated-v2
  497. 44.3 cognitivecomputations dolphin-2.6-mixtral-8x7b
  498. 44.3 freewheelin free-llama3-dpo-v0.2
  499. 44.3 Ali-C137 Q2AW1M-1001
  500. 44.3 princeton-nlp Llama-3-Instruct-8B-SLiC-HF
  501. 44.3 01-ai/Yi-Coder-9B-Chat
  502. 44.3 Ali-C137 Q2AW1M-0001
  503. 44.3 Artples L-MChat-7b
  504. 44.2 cognitivecomputations dolphin-2.7-mixtral-8x7b
  505. 44.2 ConvexAI Luminex-34B-v0.2
  506. 44.1 abacusai Smaug-34B-v0.1
  507. 44.1 EpistemeAI2 Fireball-Llama-3.1-8B-Philos-Relection
  508. 44.1 princeton-nlp Llama-3-Base-8B-SFT
  509. 44.1 mlabonne Meta-Llama-3.1-8B-Instruct-abliterated
  510. 44.1 OpenBuddy openbuddy-llama3.1-8b-v22.3-131k
  511. 44.1 princeton-nlp Llama-3-Instruct-8B-ORPO
  512. 44.0 fblgit UNA-SimpleSmaug-34b-v1beta
  513. 44.0 Omartificial-Intelligence-Space al-baka-llama3-8b-experimental
  514. 44.0 lightblue suzume-llama-3-8B-multilingual-orpo-borda-half
  515. 44.0 princeton-nlp Mistral-7B-Base-SFT-KTO
  516. 44.0 vicgalle Configurable-Mistral-7B
  517. 43.9 MaziyarPanahi Mistral-7B-Instruct-v0.3
  518. 43.9 GritLM GritLM-7B-KTO
  519. 43.9 princeton-nlp Mistral-7B-Instruct-RDPO
  520. 43.8 Qwen2-7B
  521. 43.8 ConvexAI Luminex-34B-v0.1
  522. 43.8 mistralai/Mistral-7B-Instruct-v0.3
  523. 43.8 princeton-nlp Mistral-7B-Instruct-CPO
  524. 43.8 tiiuae/falcon-180B-chat
  525. 43.8 inceptionai jais-family-1p3b-chat
  526. 43.7 princeton-nlp Llama-3-8B-ProLong-64k-Base
  527. 43.7 princeton-nlp Mistral-7B-Instruct-KTO
  528. 43.7 cognitivecomputations openchat-3.5-0106-laser
  529. 43.7 openchat openchat-3.5-0106
  530. 43.7 Ali-C137 Q2AW1M-0111
  531. 43.6 arcee-ai Virtuoso-Lite
  532. 43.6 rombodawg rombos_Replete-Coder-Instruct-8b-Merged
  533. 43.6 Ali-C137 Q2AW1M-1111
  534. 43.6 Qwen2.5-3B
  535. 43.5 cognitivecomputations dolphin-2.9.1-llama-3-8b
  536. 43.5 wenbopan Faro-Yi-34B-DPO
  537. 43.5 Mistral-7B-Instruct-v0.2
  538. 43.5 dfurman Llama-3-8B-Orpo-v0.1
  539. 43.4 failspy Meta-Llama-3-8B-Instruct-abliterated-v3
  540. 43.4 ibivibiv llama3-8b-instruct-summary-combined
  541. 43.3 Qwen2.5-Coder (7B)
  542. 43.3 lightblue suzume-llama-3-8B-multilingual
  543. 43.2 jondurbin bagel-dpo-34b-v0.5
  544. 43.1 01-ai/Yi-1.5-34B-Chat-16K
  545. 43.1 Qwen1.5-7B-Chat
  546. 43.1 INSAIT-Institute BgGPT-7B-Instruct-v0.2
  547. 43.0 Ali-C137 L3H10M-0000
  548. 43.0 ValiantLabs Llama3.1-8B-ShiningValiant2
  549. 43.0 BAAI Infinity-Instruct-3M-0625-Yi-1.5-9B
  550. 43.0 Llama 3.1-8B
  551. 43.0 princeton-nlp Mistral-7B-Instruct-IPO
  552. 42.9 meta-llama/Meta-Llama-3.1-8B
  553. 42.9 vicgalle ConfigurableHermes-7B
  554. 42.9 abacusai Llama-3-Smaug-8B
  555. 42.9 tokyotech-llm Llama-3-Swallow-8B-Instruct-v0.1
  556. 42.9 anthracite-org magnum-v3-34b
  557. 42.8 migtissera Tess-v2.5-Phi-3-medium-128k-14B
  558. 42.7 meta-llama/Llama-3.2-3B-Instruct
  559. 42.6 beowolx CodeNinja-1.0-OpenChat-7B
  560. 42.6 CombinHorizon YiSM-blossom5.1-34B-SLERP
  561. 42.6 princeton-nlp Mistral-7B-Instruct-ORPO
  562. 42.5 wenbopan Faro-Yi-9B
  563. 42.5 cognitivecomputations dolphin-2.9.1-yi-1.5-34b
  564. 42.5 SeaLLMs SeaLLM-7B-v2
  565. 42.4 BAAI Infinity-Instruct-3M-0613-Mistral-7B
  566. 42.4 altomek YiSM-34B-0rn
  567. 42.4 meta-llama/Meta-Llama-3-8B
  568. 42.3 vicgalle Configurable-Janus-7B
  569. 42.3 terrycraddock Reflection-Llama-3.1-8B
  570. 42.2 FreedomIntelligence AceGPT-7B-chat
  571. 42.2 01-ai/Yi-Coder-9B
  572. 42.2 vicgalle Configurable-Yi-1.5-9B-Chat
  573. 42.2 abacusai Liberated-Qwen1.5-14B
  574. 42.2 upstage SOLAR-10.7B-v1.0
  575. 42.1 google/gemma-7b-it
  576. 42.1 Falcon3-10B-Instruct
  577. 42.1 princeton-nlp Mistral-7B-Instruct-SLiC-HF
  578. 42.1 01-ai/Yi-1.5-34B-Chat
  579. 42.0 01-ai/Yi-1.5-9B-Chat-16K
  580. 42.0 cognitivecomputations dolphin-2.9.3-mistral-7B-32k
  581. 41.9 Qwen1.5-14B
  582. 41.9 Yi-1.5-9B-Chat
  583. 41.9 oobabooga CodeBooga-34B-v0.1
  584. 41.8 brucethemoose Yi-34B-200K-DARE-merge-v7
  585. 41.7 unsloth llama-3-8b-bnb-4bit
  586. 41.7 oddadmix Arabic-Qwen2.5-3B
  587. 41.7 wenbopan Faro-Yi-34B
  588. 41.7 chujiezheng Smaug-34B-v0.1-ExPO
  589. 41.6 meta-llama/Llama-2-70b-hf
  590. 41.5 Weyaxi Nous-Hermes-2-SUS-Chat-34B-Slerp
  591. 41.5 royallab ZephRP-m7b
  592. 41.4 byroneverson Yi-1.5-9B-Chat-16K-abliterated
  593. 41.4 byroneverson Yi-1.5-9B-Chat-abliterated
  594. 41.4 Ali-C137 Q15H10M-0000
  595. 41.4 cognitivecomputations dolphin-2.9.3-mistral-nemo-12b
  596. 41.3 sambanovasystems SambaLingo-Arabic-Chat
  597. 41.3 Falcon3-7B-Instruct
  598. 41.2 cognitivecomputations dolphin-2.9.1-yi-1.5-9b
  599. 41.1 Omartificial-Intelligence-Space Arabic-llama3.1-16bit-FT
  600. 41.1 speakleash Bielik-11B-v2.3-Instruct
  601. 41.1 rombodawg Everyone-Coder-4x7b-Base
  602. 41.0 OpenBuddy openbuddy-zero-56b-v21.2-32k
  603. 41.0 Azure99 blossom-v5.1-34b
  604. 41.0 Falcon-180B
  605. 41.0 FreedomIntelligence AceGPT-13B
  606. 41.0 qnguyen3 Master-Yi-9B
  607. 41.0 FreedomIntelligence AceGPT-v1.5-13B
  608. 41.0 oddadmix Arabic-Qwen2.5-3B-v2.1
  609. 40.9 uukuguy speechless-code-mistral-7b-v1.0
  610. 40.8 ghost-x ghost-8b-beta-1608
  611. 40.8 inceptionai jais-adapted-13b
  612. 40.8 618AI dictalm2-it-qa-fine-tune
  613. 40.7 deepseek-ai/DeepSeek-R1-Distill-Llama-8B
  614. 40.7 Qwen1.5-7B
  615. 40.6 Qwen/Qwen2.5-Math-7B
  616. 40.6 rombodawg rombos_Replete-Coder-Llama3-8B
  617. 40.4 uukuguy speechless-zephyr-code-functionary-7b
  618. 40.4 OpenBuddy openbuddy-yi1.5-34b-v21.3-32k
  619. 40.4 princeton-nlp Mistral-7B-Base-SFT-SimPO
  620. 40.4 aws-prototyping MegaBeam-Mistral-7B-512k
  621. 40.4 cognitivecomputations dolphin-2.9.2-Phi-3-Medium
  622. 40.4 SakanaAI DiscoPOP-zephyr-7b-gemma
  623. 40.3 GritLM GritLM-7B
  624. 40.3 Qwen/Qwen2.5-1.5B-Instruct
  625. 40.3 Qwen2-1.5B-Instruct
  626. 40.3 MohamedRashad Arabic-Orpo-Llama-3-8B-Instruct
  627. 40.3 speakleash Bielik-11B-v2
  628. 40.3 kaist-ai janus-dpo-7b
  629. 40.2 princeton-nlp Mistral-7B-Base-SFT-RDPO
  630. 40.2 cognitivecomputations dolphin-2.9.2-Phi-3-Medium-abliterated
  631. 40.1 TIGER-Lab MAmmoTH2-7B-Plus
  632. 40.1 kaist-ai janus-7b
  633. 40.1 cognitivecomputations dolphin-2.9.3-Yi-1.5-34B-32k
  634. 40.0 kaist-ai mistral-orpo-capybara-7k
  635. 40.0 princeton-nlp Mistral-7B-Base-SFT-SLiC-HF
  636. 40.0 uukuguy zephyr-7b-alpha-dare-0.85
  637. 39.9 cognitivecomputations dolphin-2.9.1-mixtral-1x22b
  638. 39.8 oddadmix Arabic-Qwen2.5-3B-v2
  639. 39.8 CausalLM 34b-beta
  640. 39.8 google/recurrentgemma-2b-it
  641. 39.7 amazon MegaBeam-Mistral-7B-300k
  642. 39.6 core42 jais-30b-v1
  643. 39.6 Qwen2.5-1.5B
  644. 39.6 mlabonne FineLlama-3.1-8B
  645. 39.5 01-ai/Yi-1.5-9B
  646. 39.5 NLPark AnFeng_v3.1-Avocet
  647. 39.5 01-ai/Yi-1.5-34B-32K
  648. 39.5 01-ai/Yi-1.5-9B-32K
  649. 39.5 01-ai/Yi-9B-200K
  650. 39.4 tiiuae/Falcon3-10B-Base
  651. 39.4 maywell Synatra-7B-v0.3-RP
  652. 39.4 Qwen/Qwen1.5-MoE-A2.7B
  653. 39.4 Qwen/Qwen1.5-MoE-A2.7B-Chat
  654. 39.3 princeton-nlp Mistral-7B-Base-SFT-RRHF
  655. 39.2 Ali-C137 M03H10M-0000
  656. 39.1 Sakalti Saka-1.5B
  657. 39.1 OpenBuddy openbuddy-yi1.5-34b-v21.6-32k-fp16
  658. 39.0 princeton-nlp Mistral-7B-Base-SFT-CPO
  659. 39.0 dicta-il dictalm2.0
  660. 39.0 princeton-nlp Mistral-7B-Base-SFT-IPO
  661. 38.9 Yi-1.5-34B
  662. 38.9 NotAiLOL Yi-1.5-dolphin-9B
  663. 38.8 DeepSeek-R1-Distill-Qwen-7B
  664. 38.8 Syed-Hasan-8503 Phi-3-mini-4K-instruct-cpo-simpo
  665. 38.7 Mistral 7B v0.3
  666. 38.7 MaziyarPanahi Mistral-7B-v0.3
  667. 38.6 inceptionai jais-adapted-7b
  668. 38.6 dicta-il dictalm2.0-instruct
  669. 38.6 princeton-nlp Mistral-7B-Base-SFT-DPO
  670. 38.5 sambanovasystems SambaLingo-Arabic-Base
  671. 38.3 SenseLLM ReflectionCoder-CL-34B
  672. 38.3 inceptionai jais-family-13b
  673. 38.3 kaist-ai janus-orpo-7b
  674. 38.2 BAAI OPI-Llama-3.1-8B-Instruct
  675. 38.2 migtissera Tess-M-v1.3
  676. 38.1 Gemma 1.1 2B IT
  677. 38.0 sajaw llama-2-7b-RandomGPT-5K-ar
  678. 38.0 FreedomIntelligence AceGPT-7B
  679. 38.0 abacusai bigyi-15b
  680. 38.0 tiiuae/Falcon3-7B-Base
  681. 37.9 core42 jais-13b
  682. 37.9 haoranxu ALMA-13B-Pretrain
  683. 37.9 Qwen2-1.5B
  684. 37.8 google/gemma-2b-it
  685. 37.7 Qwen/Qwen1.5-4B
  686. 37.6 meta-llama/Llama-2-13b-hf
  687. 37.6 gpt2
  688. 37.6 inceptionai jais-family-6p7b
  689. 37.5 Qwen/Qwen1.5-1.8B-Chat
  690. 37.5 Qwen/Qwen2.5-Math-7B-Instruct
  691. 37.4 google/recurrentgemma-2b
  692. 37.3 Qwen1.5-4B-Chat
  693. 37.3 Ali-C137 L2H10M-0000
  694. 37.2 meta-llama/Llama-2-7b-hf
  695. 37.0 tiiuae/Falcon3-1B-Base
  696. 37.0 SenseLLM ReflectionCoder-DS-33B
  697. 36.8 inceptionai jais-family-590m-chat
  698. 36.7 Qwen/Qwen1.5-1.8B
  699. 36.6 haoranxu ALMA-13B-R
  700. 36.5 Qwen2-0.5B-Instruct
  701. 36.5 byroneverson Yi-1.5-34B-Chat-abliterated
  702. 36.5 NTQAI Nxcode-CQ-7B-orpo
  703. 36.4 Qwen2.5-0.5B
  704. 36.3 inceptionai jais-family-2p7b
  705. 36.2 Qwen/Qwen1.5-0.5B-Chat
  706. 36.2 Llama 3.2 1B
  707. 36.1 haoranxu ALMA-7B-R
  708. 36.1 Falcon3-3B-Instruct
  709. 36.0 Falcon3-1B-Instruct
  710. 36.0 inceptionai jais-family-1p3b
  711. 36.0 Gemma 2B
  712. 36.0 riotu-lab ArabianGPT-0.3B-QA
  713. 35.9 inceptionai jais-family-590m
  714. 35.9 bigscience bloom-7b1
  715. 35.7 tiiuae/falcon-11B
  716. 35.7 Qwen/Qwen1.5-0.5B
  717. 35.7 riotu-lab ArabianGPT-01B
  718. 35.5 riotu-lab ArabianGPT-03B
  719. 35.5 tiiuae/falcon-7b-instruct
  720. 35.5 Falcon-40B
  721. 35.4 riotu-lab ArabianGPT-08B-V2
  722. 35.4 kevinpro Hydra-LLaMA3-8B-v0.1
  723. 35.3 Qwen2-0.5B
  724. 35.2 Artples L-MChat-Small
  725. 35.2 tiiuae/Falcon3-3B-Base
  726. 35.1 kevinpro Hydra-LLaMA3-8B-v0.2
  727. 35.0 Falcon-7B
  728. 35.0 Ali-C137 F1H10M-0000
  729. 35.0 paloalma ECE-TW3-JRGL-V3
  730. 34.8 paloalma ECE-TW3-JRGL-V4
  731. 34.7 riotu-lab ArabianGPT-1.5B
  732. 34.6 tiiuae/falcon-40b-instruct
  733. 34.5 kaist-ai janus-rm-7b
  734. 34.4 MBZUAI MobiLlama-1B
  1. 1 silma-ai SILMA-9B-Instruct-v1.0 chat models (RLHF, DPO, IFT, ...); bfloat16 70.6%
  2. 2 MaziyarPanahi calme-2.1-qwen2.5-72b chat models (RLHF, DPO, IFT, ...); bfloat16 69.2%
  3. 3 Sakalti ultiima-72B fine-tuned on domain-specific datasets; float16 69.1%
  4. 4 Replete-AI Replete-LLM-V2.5-Qwen-72b fine-tuned on domain-specific datasets; bfloat16 69.1%
  5. 5 MaziyarPanahi calme-2.2-qwen2.5-72b chat models (RLHF, DPO, IFT, ...); bfloat16 69.1%
  6. 6 rombodawg Rombos-LLM-V2.5-Qwen-72b base merges and moerges; bfloat16 69.1%
  7. 7 Qwen2.5-72B chat models (RLHF, DPO, IFT, ...); bfloat16 69.1%
  8. 8 zetasepic Qwen2.5-72B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 68.1%
  9. 9 paloalma TW3-JRGL-v2 base merges and moerges; bfloat16 67.1%
  10. 10 davidkim205 Rhea-72b-v0.5 chat models (RLHF, DPO, IFT, ...); bfloat16 67%
  11. 11 AXCXEPT EZO-Qwen2.5-72B-Instruct fine-tuned on domain-specific datasets; bfloat16 66.5%
  12. 12 freewheelin free-evo-qwen72b-v0.8-re fine-tuned on domain-specific datasets; float16 65.9%
  13. 13 MaziyarPanahi calme-2.2-qwen2-72b chat models (RLHF, DPO, IFT, ...); bfloat16 65.8%
  14. 14 dnhkng RYS-XLarge chat models (RLHF, DPO, IFT, ...); bfloat16 65.1%
  15. 15 arcee-ai Arcee-Nova chat models (RLHF, DPO, IFT, ...); bfloat16 65.1%
  16. 16 dnhkng RYS-XLarge-base fine-tuned on domain-specific datasets; bfloat16 65%
  17. 17 MTSAIR MultiVerse_70B fine-tuned on domain-specific datasets; bfloat16 65%
  18. 18 anthracite-org magnum-v1-72b fine-tuned on domain-specific datasets; bfloat16 64.7%
  19. 19 Undi95 MG-FinalMix-72B fine-tuned on domain-specific datasets; bfloat16 64.7%
  20. 20 MaziyarPanahi calme-2.1-qwen2-72b fine-tuned on domain-specific datasets; bfloat16 64.7%
  21. 21 dfurman Qwen2-72B-Orpo-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 64.7%
  22. 22 MaziyarPanahi Qwen2-72B-Instruct-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 64.7%
  23. 23 alpindale magnum-72b-v1 fine-tuned on domain-specific datasets; bfloat16 64.7%
  24. 24 abacusai Dracarys-72B-Instruct fine-tuned on domain-specific datasets; bfloat16 64.6%
  25. 25 silma-ai SILMA-9B-Instruct-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 64.4%
  26. 26 Nexusflow Athene-70B ; bfloat16 64.4%
  27. 27 mssqpi G3 chat models (RLHF, DPO, IFT, ...); bfloat16 64.4%
  28. 28 silma-ai SILMA-9B-Instruct-v0.1.0 chat models (RLHF, DPO, IFT, ...); bfloat16 64.3%
  29. 29 silma-ai SILMA-9B-Instruct-v0.1.1 pretrained; bfloat16 64.3%
  30. 30 tanliboy lambda-qwen2.5-32b-dpo-test chat models (RLHF, DPO, IFT, ...); bfloat16 64%
  31. 31 Sakalti ultiima-125B fine-tuned on domain-specific datasets; float16 64%
  32. 32 anthracite-org magnum-v2-72b chat models (RLHF, DPO, IFT, ...); bfloat16 63.7%
  33. 33 Qwen2-72B chat models (RLHF, DPO, IFT, ...); bfloat16 63.6%
  34. 34 paloalma Le_Triomphant-ECE-TW3 base merges and moerges; bfloat16 63.5%
  35. 35 abacusai Smaug-Qwen2-72B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 63.4%
  36. 36 pankajmathur orca_mini_v7_72b chat models (RLHF, DPO, IFT, ...); bfloat16 63.1%
  37. 37 Qwen/Qwen2.5-32B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 63.1%
  38. 38 paulml ECE-ILAB-Q1 base merges and moerges; bfloat16 63%
  39. 39 Sakalti ultiima-32B base merges and moerges; float16 63%
  40. 40 Replete-AI Replete-LLM-V2.5-Qwen-32b fine-tuned on domain-specific datasets; bfloat16 62.9%
  41. 41 VAGOsolutions Llama-3-SauerkrautLM-70b-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 62.8%
  42. 42 Sakalti ultiima-78B fine-tuned on domain-specific datasets; float16 62.7%
  43. 43 dfurman CalmeRys-78B-Orpo-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 62.6%
  44. 44 sometimesanotion Qwen2.5-14B-Vimarckoso-v3 base merges and moerges; bfloat16 62.4%
  45. 45 sometimesanotion Lamarck-14B-v0.6 base merges and moerges; bfloat16 62.3%
  46. 46 Sakalti ultiima-78B-V2 base merges and moerges; float16 62.3%
  47. 47 hotmailuser QwenSlerp2-14B base merges and moerges; bfloat16 62.3%
  48. 48 AXCXEPT EZO-Qwen2.5-32B-Instruct fine-tuned on domain-specific datasets; bfloat16 62.2%
  49. 49 Sakalti SakaMoe-3x14B-Instruct fine-tuned on domain-specific datasets; float16 62.1%
  50. 50 tanliboy lambda-qwen2.5-14b-dpo-test chat models (RLHF, DPO, IFT, ...); bfloat16 62%
  51. 51 huihui-ai Qwen2.5-32B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 61.8%
  52. 52 sometimesanotion Lamarck-14B-v0.7 chat models (RLHF, DPO, IFT, ...); bfloat16 61.7%
  53. 53 Slim205 Barka-9b-it-v02 fine-tuned on domain-specific datasets; bfloat16 61.7%
  54. 54 airev-ai Amal-70b-v2 fine-tuned on domain-specific datasets; float16 61.7%
  55. 55 airev-ai Amal-70b-v1.1 ; float16 61.6%
  56. 56 migtissera Tess-v2.5.2-Qwen2-72B fine-tuned on domain-specific datasets; bfloat16 61.6%
  57. 57 airev-ai Amal-75b-v3 ; float16 61.5%
  58. 58 byroneverson gemma-2-27b-it-abliterated fine-tuned on domain-specific datasets; bfloat16 61.5%
  59. 59 MaziyarPanahi calme-2.4-llama3-70b chat models (RLHF, DPO, IFT, ...); bfloat16 61.4%
  60. 60 Lambent qwen2.5-reinstruct-alternate-lumen-14B base merges and moerges; bfloat16 61.3%
  61. 61 v000000 Qwen2.5-Lumen-14B chat models (RLHF, DPO, IFT, ...); bfloat16 61.3%
  62. 62 MaziyarPanahi calme-2.2-llama3-70b chat models (RLHF, DPO, IFT, ...); bfloat16 61.2%
  63. 63 rombodawg Rombos-LLM-V2.6-Qwen-14b fine-tuned on domain-specific datasets; bfloat16 61.2%
  64. 64 Replete-AI Replete-LLM-V2.5-Qwen-14b fine-tuned on domain-specific datasets; bfloat16 61.1%
  65. 65 Sakalti ultiima-14B base merges and moerges; float16 61.1%
  66. 66 v000000 Qwen2.5-14B-Gutenberg-1e-Delta chat models (RLHF, DPO, IFT, ...); bfloat16 61.1%
  67. 67 airev-ai qwen-14b-v1 ; float16 61.1%
  68. 68 Qwen/Qwen2.5-14B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 61.1%
  69. 69 v000000 Qwen2.5-14B-Gutenberg-Instruct-Slerpeno base merges and moerges; bfloat16 61%
  70. 70 VAGOsolutions Llama-3.1-SauerkrautLM-70b-Instruct fine-tuned on domain-specific datasets; bfloat16 60.9%
  71. 71 FiditeNemini Unhinged-Qwen2-72B base merges and moerges; bfloat16 60.9%
  72. 72 yellowtown 7B-v0.2 fine-tuned on domain-specific datasets; bfloat16 60.9%
  73. 73 MaziyarPanahi calme-2.3-llama3-70b chat models (RLHF, DPO, IFT, ...); bfloat16 60.8%
  74. 74 NousResearch Hermes-3-Llama-3.1-70B chat models (RLHF, DPO, IFT, ...); bfloat16 60.8%
  75. 75 google/gemma-2-27b-it chat models (RLHF, DPO, IFT, ...); bfloat16 60.6%
  76. 76 elmrc JuhainaG2 chat models (RLHF, DPO, IFT, ...); bfloat16 60.2%
  77. 77 MaziyarPanahi Llama-3-70B-Instruct-DPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 60.1%
  78. 78 mmnga Llama-3-70B-japanese-suzume-vector-v0.1 fine-tuned on domain-specific datasets; bfloat16 60%
  79. 79 airev-ai Jais-Inception-70b-V1.2.2 ; float16 59.9%
  80. 80 meta-llama/Meta-Llama-3-70B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 59.9%
  81. 81 failspy llama-3-70B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 59.8%
  82. 82 dnhkng RYS-Gemma-2-27b-it fine-tuned on domain-specific datasets; bfloat16 59.5%
  83. 83 airev-ai qwen-14b-v2 ; float16 59.5%
  84. 84 airev-ai Jais-Inception-70b-V1.1 chat models (RLHF, DPO, IFT, ...); bfloat16 59.4%
  85. 85 recoilme recoilme-gemma-2-9B-v0.4 fine-tuned on domain-specific datasets; bfloat16 59.3%
  86. 86 cognitivecomputations dolphin-2.9.2-qwen2-72b chat models (RLHF, DPO, IFT, ...); bfloat16 59.1%
  87. 87 migtissera Tess-v2.5-Qwen2-72B fine-tuned on domain-specific datasets; float16 59%
  88. 88 Qwen2.5-14B-Instruct-1M chat models (RLHF, DPO, IFT, ...); bfloat16 58.8%
  89. 89 abacusai Smaug-72B-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 58.8%
  90. 90 elmrc juhainaG fine-tuned on domain-specific datasets; bfloat16 58.6%
  91. 91 princeton-nlp gemma-2-9b-it-SimPO chat models (RLHF, DPO, IFT, ...); bfloat16 58.5%
  92. 92 elmresearchcenter juhaina fine-tuned on domain-specific datasets; bfloat16 58.5%
  93. 93 gbueno86 Meta-LLama-3-Cat-Smaug-LLama-70b base merges and moerges; bfloat16 58.4%
  94. 94 Nohobby Q2.5-Qwetiapin-32B base merges and moerges; bfloat16 58.2%
  95. 95 anthracite-org magnum-v3-27b-kto ; ? 58.2%
  96. 96 nvidia Llama-3.1-Nemotron-70B-Instruct-HF chat models (RLHF, DPO, IFT, ...); bfloat16 58%
  97. 97 Daemontatox Mawared_GS fine-tuned on domain-specific datasets; bfloat16 58%
  98. 98 airev-ai Amal-70b-v5 ; float16 58%
  99. 99 arcee-ai Virtuoso-Small chat models (RLHF, DPO, IFT, ...); bfloat16 57.9%
  100. 100 maldv Qwenstein2.5-32B-Instruct base merges and moerges; bfloat16 57.8%
  101. 101 abhishek autotrain-llama3-70b-orpo-v1 chat models (RLHF, DPO, IFT, ...); float16 57.8%
  102. 102 gbueno86 Meta-LLama-3-Cat-A-LLama-70b base merges and moerges; bfloat16 57.6%
  103. 103 abhishek autotrain-llama3-70b-orpo-v2 chat models (RLHF, DPO, IFT, ...); float16 57.5%
  104. 104 VAGOsolutions SauerkrautLM-Nemo-12b-Instruct fine-tuned on domain-specific datasets; bfloat16 57.4%
  105. 105 UCLA-AGI Gemma-2-9B-It-SPPO-Iter3 chat models (RLHF, DPO, IFT, ...); bfloat16 57.2%
  106. 106 MaziyarPanahi calme-2.3-llama3.1-70b chat models (RLHF, DPO, IFT, ...); bfloat16 57.1%
  107. 107 UCLA-AGI Gemma-2-9B-It-SPPO-Iter1 chat models (RLHF, DPO, IFT, ...); bfloat16 57%
  108. 108 airev-ai emirati-14b-v2 fine-tuned on domain-specific datasets; bfloat16 57%
  109. 109 google/gemma-2-9b-it chat models (RLHF, DPO, IFT, ...); bfloat16 57%
  110. 110 Ashmal MBZUAI-ORYX-new ; float16 57%
  111. 111 airev-ai Jais-Inception-70b-V1.2 ; float16 57%
  112. 112 recoilme Gemma-2-Ataraxy-Gemmasutra-9B-slerp base merges and moerges; bfloat16 57%
  113. 113 UCLA-AGI Gemma-2-9B-It-SPPO-Iter2 chat models (RLHF, DPO, IFT, ...); bfloat16 57%
  114. 114 MaziyarPanahi calme-2.2-llama3.1-70b chat models (RLHF, DPO, IFT, ...); bfloat16 56.9%
  115. 115 cognitivecomputations dolphin-2.9.1-llama-3-70b fine-tuned on domain-specific datasets; bfloat16 56.9%
  116. 116 rombodawg Rombos-LLM-V2.6-Nemotron-70b base merges and moerges; bfloat16 56.9%
  117. 117 inceptionai jais-adapted-70b-chat ; ? 56.8%
  118. 118 elmrc JuhainaG9B chat models (RLHF, DPO, IFT, ...); bfloat16 56.8%
  119. 119 failspy Llama-3-70B-Instruct-abliterated-v3 fine-tuned on domain-specific datasets; bfloat16 56.8%
  120. 120 shenzhi-wang Llama3-70B-Chinese-Chat chat models (RLHF, DPO, IFT, ...); bfloat16 56.8%
  121. 121 mssqpi Waie1 chat models (RLHF, DPO, IFT, ...); bfloat16 56.8%
  122. 122 TareksLab Thalassic-Alpha-LLaMa-70B base merges and moerges; bfloat16 56.8%
  123. 123 airev-ai Amal-70b-v4 ; float16 56.8%
  124. 124 hotmailuser Deeepseek-QwenSlerp4-32B base merges and moerges; bfloat16 56.6%
  125. 125 sambanovasystems SambaLingo-Arabic-Chat-70B ; ? 56.6%
  126. 126 recoilme recoilme-gemma-2-9B-v0.2 fine-tuned on domain-specific datasets; bfloat16 56.5%
  127. 127 Steelskull L3-MS-Astoria-70b base merges and moerges; bfloat16 56.4%
  128. 128 MaziyarPanahi calme-2.3-qwen2-72b chat models (RLHF, DPO, IFT, ...); bfloat16 56.4%
  129. 129 CohereForAI/c4ai-command-r-plus fine-tuned on domain-specific datasets; float16 56.4%
  130. 130 meta-llama/Meta-Llama-3.1-70B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 56.2%
  131. 131 tenyx Llama3-TenyxChat-70B chat models (RLHF, DPO, IFT, ...); bfloat16 56.2%
  132. 132 BAAI Gemma2-9B-IT-Simpo-Infinity-Preference chat models (RLHF, DPO, IFT, ...); bfloat16 55.9%
  133. 133 inceptionai jais-family-30b-8k-chat ; ? 55.9%
  134. 134 Ashmal MBZUAI-oryx fine-tuned on domain-specific datasets; float16 55.8%
  135. 135 migtissera Llama-3-70B-Synthia-v3.5 fine-tuned on domain-specific datasets; float16 55.7%
  136. 136 CohereForAI/c4ai-command-r-plus-08-2024 chat models (RLHF, DPO, IFT, ...); float16 55.7%
  137. 137 princeton-nlp gemma-2-9b-it-DPO ; bfloat16 55.7%
  138. 138 recoilme recoilme-gemma-2-9B-v0.3 fine-tuned on domain-specific datasets; bfloat16 55.5%
  139. 139 Triangle104 DeepSky-T100 base merges and moerges; float16 55.3%
  140. 140 airev-ai emirati-14b-v3 fine-tuned on domain-specific datasets; bfloat16 55.2%
  141. 141 Gemma 2 27B ; ? 55.2%
  142. 142 huihui-ai Qwen2.5-7B-Instruct-abliterated-v2 fine-tuned on domain-specific datasets; bfloat16 55%
  143. 143 Delta-Vector Odin-9B fine-tuned on domain-specific datasets; bfloat16 55%
  144. 144 FreedomIntelligence AceGPT-v2-70B pretrained; float16 55%
  145. 145 prithivMLmods Qwen2.5-32B-DeepSeek-R1-Instruct base merges and moerges; bfloat16 54.9%
  146. 146 Joseph717171 Llama-3.1-SuperNova-8B-Lite_TIES_with_Base base merges and moerges; bfloat16 54.9%
  147. 147 Ba2han QwQenSeek-coder base merges and moerges; bfloat16 54.8%
  148. 148 FreedomIntelligence AceGPT-v2-32B-Chat chat models (RLHF, DPO, IFT, ...); float16 54.6%
  149. 149 core42 jais-30b-chat-v3 ; ? 54.5%
  150. 150 Orion-zhen Qwen2.5-7B-Instruct-Uncensored chat models (RLHF, DPO, IFT, ...); bfloat16 54.5%
  151. 151 anthracite-org magnum-v4-9b chat models (RLHF, DPO, IFT, ...); bfloat16 54.5%
  152. 152 Command R (03-2024) fine-tuned on domain-specific datasets; float16 54.4%
  153. 153 Isaak-Carter Josiefied-Qwen2.5-7B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 54.3%
  154. 154 BAAI Infinity-Instruct-7M-Gen-Llama3_1-70B fine-tuned on domain-specific datasets; bfloat16 54.3%
  155. 155 Qwen/Qwen2.5-7B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 54.3%
  156. 156 Qwen2.5-32B pretrained; bfloat16 54.2%
  157. 157 ai21labs AI21-Jamba-1.5-Mini pretrained; bfloat16 54.2%
  158. 158 anthracite-org magnum-v3-9b-customgemma2 chat models (RLHF, DPO, IFT, ...); bfloat16 54.2%
  159. 159 Replete-AI Replete-LLM-V2.5-Qwen-7b fine-tuned on domain-specific datasets; bfloat16 54.1%
  160. 160 TheDrummer Gemmasutra-9B-v1 base merges and moerges; bfloat16 54.1%
  161. 161 migtissera Tess-2.0-Llama-3-70B fine-tuned on domain-specific datasets; float16 54.1%
  162. 162 jondurbin airoboros-dpo-70b-3.3 chat models (RLHF, DPO, IFT, ...); bfloat16 54%
  163. 163 failspy Meta-Llama-3-70B-Instruct-abliterated-v3.5 fine-tuned on domain-specific datasets; bfloat16 54%
  164. 164 migtissera Tess-2.0-Llama-3-70B-v0.2 fine-tuned on domain-specific datasets; float16 53.9%
  165. 165 jondurbin airoboros-70b-3.3 chat models (RLHF, DPO, IFT, ...); bfloat16 53.8%
  166. 166 Sakalti Saka-7.6B base merges and moerges; float16 53.8%
  167. 167 RDson WomboCombo-R1-Coder-14B-Preview base merges and moerges; bfloat16 53.8%
  168. 168 c4ai-command-r-08-2024 chat models (RLHF, DPO, IFT, ...); float16 53.7%
  169. 169 mistralai/Codestral-22B-v0.1 fine-tuned on domain-specific datasets; bfloat16 53.7%
  170. 170 airev-ai Jais-Inception-70b ; bfloat16 53.7%
  171. 171 DeepSeek-R1-Distill-Llama-70B fine-tuned on domain-specific datasets; bfloat16 53.6%
  172. 172 arcee-ai Arcee-Spark chat models (RLHF, DPO, IFT, ...); bfloat16 53.6%
  173. 173 Daemontatox Llama3.3-70B-CogniLink fine-tuned on domain-specific datasets; bfloat16 53.5%
  174. 174 huihui-ai Qwen2.5-7B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 53.5%
  175. 175 mssqpi ElmGemma2 chat models (RLHF, DPO, IFT, ...); bfloat16 53.5%
  176. 176 Isaak-Carter Josiefied-Qwen2.5-7B-Instruct-abliterated-v2 fine-tuned on domain-specific datasets; bfloat16 53.4%
  177. 177 Qwen2.5-14B pretrained; bfloat16 53.4%
  178. 178 Cran-May T.E-8.1 chat models (RLHF, DPO, IFT, ...); bfloat16 53.3%
  179. 179 dfurman Llama-3-70B-Orpo-v0.1 chat models (RLHF, DPO, IFT, ...); float16 53.3%
  180. 180 Aya Expanse 8B chat models (RLHF, DPO, IFT, ...); float16 53.3%
  181. 181 migtissera Tess-3-Llama-3.1-70B fine-tuned on domain-specific datasets; bfloat16 53.1%
  182. 182 Qwen2.5-Math-72B-Instruct fine-tuned on domain-specific datasets; bfloat16 53.1%
  183. 183 paloalma ECE-TW3-JRGL-V2 base merges and moerges; bfloat16 53.1%
  184. 184 dnhkng RYS-Llama3.1-Large fine-tuned on domain-specific datasets; bfloat16 53.1%
  185. 185 vicgalle CarbonBeagle-11B base merges and moerges; float16 53%
  186. 186 inceptionai jais-family-13b-chat chat models (RLHF, DPO, IFT, ...); float16 52.9%
  187. 187 Qwen/Qwen2-MoE-57B-A14B-Instruct ; bfloat16 52.9%
  188. 188 SicariusSicariiStuff Qwen2.5-14B_Uncencored fine-tuned on domain-specific datasets; float16 52.8%
  189. 189 SeaLLMs SeaLLM-7B-v2.5 fine-tuned on domain-specific datasets; bfloat16 52.7%
  190. 190 meta-llama/Meta-Llama-3-70B pretrained; bfloat16 52.7%
  191. 191 nitky RoguePlanet-DeepSeek-R1-Qwen-32B base merges and moerges; bfloat16 52.6%
  192. 192 meta-llama/Meta-Llama-3.1-70B pretrained; bfloat16 52.6%
  193. 193 anthracite-org magnum-v3-9b-chatml chat models (RLHF, DPO, IFT, ...); bfloat16 52.5%
  194. 194 Llama 3.1-70B pretrained; bfloat16 52.5%
  195. 195 abacusai Smaug-Llama-3-70B-Instruct-32K chat models (RLHF, DPO, IFT, ...); bfloat16 52.4%
  196. 196 FreedomIntelligence AceGPT-v2-70B-Chat chat models (RLHF, DPO, IFT, ...); float16 52.4%
  197. 197 informatiker Qwen2-7B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 52.4%
  198. 198 CohereForAI/aya-23-35B fine-tuned on domain-specific datasets; float16 52.3%
  199. 199 TeetouchQQ model_mergev2 base merges and moerges; bfloat16 52.3%
  200. 200 abacusai Smaug-Llama-3-70B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 52.2%
  201. 201 natong19 Qwen2-7B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 52.2%
  202. 202 xiangxinai Xiangxin-2XL-Chat-1048k-Chinese-Llama3-70B chat models (RLHF, DPO, IFT, ...); float16 52%
  203. 203 OpenBuddy openbuddy-llama3-70b-v21.2-32k chat models (RLHF, DPO, IFT, ...); bfloat16 52%
  204. 204 SicariusSicariiStuff Qwen2.5-14B_Uncensored_Instruct chat models (RLHF, DPO, IFT, ...); float16 51.8%
  205. 205 failspy Smaug-Llama-3-70B-Instruct-abliterated-v3 fine-tuned on domain-specific datasets; bfloat16 51.8%
  206. 206 vicgalle ConfigurableBeagle-11B chat models (RLHF, DPO, IFT, ...); float16 51.7%
  207. 207 Qwen2-7B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 51.7%
  208. 208 DeepMount00 Llama-3.1-8b-Ita fine-tuned on domain-specific datasets; bfloat16 51.7%
  209. 209 arcee-ai Llama-3.1-SuperNova-Lite fine-tuned on domain-specific datasets; bfloat16 51.7%
  210. 210 nyunai nyun-c2-llama3-56B fine-tuned on domain-specific datasets; float16 51.6%
  211. 211 princeton-nlp Llama-3-Base-8B-SFT-SimPO chat models (RLHF, DPO, IFT, ...); bfloat16 51.5%
  212. 212 paloalma ECE-TW3-JRGL-V1 base merges and moerges; float16 51.4%
  213. 213 VAGOsolutions Llama-3.1-SauerkrautLM-8b-Instruct fine-tuned on domain-specific datasets; bfloat16 51.4%
  214. 214 Qwen/Qwen2.5-Math-72B pretrained; bfloat16 51.4%
  215. 215 vicgalle CarbonBeagle-11B-truthy fine-tuned on domain-specific datasets; float16 51.3%
  216. 216 NaniDAO deepseek-r1-qwen-2.5-32B-ablated fine-tuned on domain-specific datasets; bfloat16 51.3%
  217. 217 utter-project EuroLLM-9B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 51.2%
  218. 218 elmrc ElmJuhaina70B chat models (RLHF, DPO, IFT, ...); bfloat16 51.2%
  219. 219 core42 jais-30b-chat-v1 ; ? 51.1%
  220. 220 nbeerbower DeepSeek-R1-Qwen-lorablated-32B base merges and moerges; bfloat16 51%
  221. 221 princeton-nlp Llama-3-Base-8B-SFT-IPO chat models (RLHF, DPO, IFT, ...); bfloat16 50.9%
  222. 222 FreedomIntelligence AceGPT-v2-32B ; float16 50.8%
  223. 223 nicoboss DeepSeek-R1-Distill-Qwen-32B-Uncensored fine-tuned on domain-specific datasets; bfloat16 50.8%
  224. 224 shadowml Mixolar-4x7b base merges and moerges; float16 50.8%
  225. 225 asas-ai AceGPT-v2-32B fine-tuned on domain-specific datasets; float16 50.8%
  226. 226 axolotl-ai-co romulus-mistral-nemo-12b-simpo chat models (RLHF, DPO, IFT, ...); bfloat16 50.8%
  227. 227 Qwen1.5-110B pretrained; bfloat16 50.6%
  228. 228 invalid-coder Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp base merges and moerges; float16 50.5%
  229. 229 jeonsworld CarbonVillain-en-10.7B-v4 base merges and moerges; float16 50.4%
  230. 230 kekmodel StopCarbon-10.7B-v5 base merges and moerges; float16 50.4%
  231. 231 Gemma 2 9B ; ? 50.4%
  232. 232 nbeerbower mistral-nemo-wissenschaft-12B chat models (RLHF, DPO, IFT, ...); bfloat16 50.3%
  233. 233 FreedomIntelligence AceGPT-v1.5-13B-Chat ; ? 50.3%
  234. 234 microsoft/Phi-3.5-mini-instruct ; bfloat16 50.2%
  235. 235 FreedomIntelligence AceGPT-v2-8B-Chat chat models (RLHF, DPO, IFT, ...); float16 50.2%
  236. 236 Sakalti Saka-3.8B base merges and moerges; float16 50.1%
  237. 237 openchat openchat-3.5-0106-gemma chat models (RLHF, DPO, IFT, ...); bfloat16 50%
  238. 238 prithivMLmods Qwen2.5-14B-DeepSeek-R1-1M base merges and moerges; bfloat16 50%
  239. 239 upstage SOLAR-10.7B-Instruct-v1.0 fine-tuned on domain-specific datasets; float16 50%
  240. 240 vicgalle ConfigurableSOLAR-10.7B chat models (RLHF, DPO, IFT, ...); float16 50%
  241. 241 migtissera Tess-v2.5-Gemma-2-27B-alpha fine-tuned on domain-specific datasets; bfloat16 49.9%
  242. 242 RDson CoderO1-DeepSeekR1-Coder-32B-Preview base merges and moerges; bfloat16 49.9%
  243. 243 ai21labs Jamba-v0.1 pretrained; bfloat16 49.9%
  244. 244 anthracite-org magnum-v2.5-12b-kto chat models (RLHF, DPO, IFT, ...); float16 49.9%
  245. 245 Sakalti Saka-24B base merges and moerges; float16 49.9%
  246. 246 NLPark Shi-Cis-Kestrel-uncensored chat models (RLHF, DPO, IFT, ...); bfloat16 49.8%
  247. 247 vicgalle Configurable-Hermes-2-Pro-Llama-3-8B chat models (RLHF, DPO, IFT, ...); float16 49.8%
  248. 248 llm-mle ara_mle_pretrained_8b ; bfloat16 49.8%
  249. 249 llm-mle arb_mle_pretrained_8b pretrained; bfloat16 49.8%
  250. 250 arcee-ai Llama-Spark ; ? 49.7%
  251. 251 princeton-nlp Llama-3-Base-8B-SFT-RDPO chat models (RLHF, DPO, IFT, ...); bfloat16 49.7%
  252. 252 MarinaraSpaghetti NemoReRemix-12B base merges and moerges; bfloat16 49.7%
  253. 253 Qwen/Qwen1.5-110B-Chat pretrained; bfloat16 49.7%
  254. 254 AALF gemma-2-27b-it-SimPO-37K-100steps chat models (RLHF, DPO, IFT, ...); bfloat16 49.7%
  255. 255 princeton-nlp Llama-3-Base-8B-SFT-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 49.6%
  256. 256 Magpie-Align Llama-3-8B-Magpie-Align-v0.3 chat models (RLHF, DPO, IFT, ...); bfloat16 49.4%
  257. 257 grimjim Llama-3-Instruct-8B-SPPO-Iter3-SimPO-merge base merges and moerges; bfloat16 49.3%
  258. 258 MaziyarPanahi calme-2.3-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 49.3%
  259. 259 AALF gemma-2-27b-it-SimPO-37K chat models (RLHF, DPO, IFT, ...); bfloat16 49.2%
  260. 260 vilm Quyen-Pro-Max-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 49.2%
  261. 261 Dampfinchen Llama-3.1-8B-Ultra-Instruct base merges and moerges; bfloat16 49.1%
  262. 262 openchat openchat-3.6-8b-20240522 chat models (RLHF, DPO, IFT, ...); bfloat16 49.1%
  263. 263 Daemontatox MawaredT1 ; bfloat16 49.1%
  264. 264 core42 jais-13b-chat ; ? 49%
  265. 265 saltlux luxia-21.4b-alignment-v1.2 chat models (RLHF, DPO, IFT, ...); bfloat16 49%
  266. 266 RLHFlow LLaMA3-iterative-DPO-final chat models (RLHF, DPO, IFT, ...); bfloat16 49%
  267. 267 turboderp Cat-Llama-3-70B-instruct chat models (RLHF, DPO, IFT, ...); float16 49%
  268. 268 inceptionai jais-adapted-70b chat models (RLHF, DPO, IFT, ...); float16 49%
  269. 269 Eurdem Defne-llama3.1-8B base merges and moerges; bfloat16 49%
  270. 270 NousResearch Nous-Hermes-2-Mixtral-8x7B-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 48.9%
  271. 271 cognitivecomputations WestLake-7B-v2-laser ; float16 48.9%
  272. 272 natong19 Mistral-Nemo-Instruct-2407-abliterated fine-tuned on domain-specific datasets; bfloat16 48.9%
  273. 273 elmrc juhaina fine-tuned on domain-specific datasets; bfloat16 48.8%
  274. 274 elmrc Juhaina28B ; ? 48.8%
  275. 275 lightblue suzume-llama-3-8B-multilingual-orpo-borda-top75 chat models (RLHF, DPO, IFT, ...); bfloat16 48.8%
  276. 276 huihui-ai DeepSeek-R1-Distill-Qwen-32B-abliterated fine-tuned on domain-specific datasets; bfloat16 48.7%
  277. 277 VAGOsolutions SauerkrautLM-Phi-3-medium chat models (RLHF, DPO, IFT, ...); bfloat16 48.7%
  278. 278 VAGOsolutions SauerkrautLM-Qwen-32b chat models (RLHF, DPO, IFT, ...); bfloat16 48.7%
  279. 279 princeton-nlp Llama-3-Instruct-8B-SimPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 48.7%
  280. 280 Qwen2.5-7B pretrained; bfloat16 48.6%
  281. 281 lemon-mint gemma-ko-7b-instruct-v0.62 fine-tuned on domain-specific datasets; bfloat16 48.6%
  282. 282 princeton-nlp Llama-3-Base-8B-SFT-KTO chat models (RLHF, DPO, IFT, ...); bfloat16 48.6%
  283. 283 Danielbrdz Barcenas-14b-Phi-3-medium-ORPO chat models (RLHF, DPO, IFT, ...); float16 48.5%
  284. 284 elmrc Juhaina29B chat models (RLHF, DPO, IFT, ...); bfloat16 48.5%
  285. 285 Mihaiii Llama-3-pruned-45B-Drobeta-Turnu-Severin ; bfloat16 48.5%
  286. 286 princeton-nlp Llama-3-8B-ProLong-64k-Instruct ; ? 48.5%
  287. 287 v000000 L3.1-Niitorm-8B-DPO-t0.0001 chat models (RLHF, DPO, IFT, ...); float16 48.5%
  288. 288 MaziyarPanahi Llama-3-8B-Instruct-v0.10 chat models (RLHF, DPO, IFT, ...); bfloat16 48.4%
  289. 289 princeton-nlp Llama-3-8B-ProLong-512k-Instruct ; ? 48.3%
  290. 290 TIGER-Lab MAmmoTH2-8x7B-Plus fine-tuned on domain-specific datasets; bfloat16 48.3%
  291. 291 MaziyarPanahi calme-2.5-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 48.3%
  292. 292 Qwen1.5-72B pretrained; bfloat16 48.3%
  293. 293 jondurbin bagel-8b-v1.0 fine-tuned on domain-specific datasets; bfloat16 48.2%
  294. 294 GritLM GritLM-8x7B-KTO chat models (RLHF, DPO, IFT, ...); bfloat16 48.2%
  295. 295 inceptionai jais-family-6p7b-chat ; ? 48.2%
  296. 296 CohereForAI/aya-23-8B fine-tuned on domain-specific datasets; float16 48.1%
  297. 297 mlabonne NeuralDaredevil-8B-abliterated chat models (RLHF, DPO, IFT, ...); float16 48%
  298. 298 MaziyarPanahi Calme-4x7B-MoE-v0.1 base merges and moerges; bfloat16 48%
  299. 299 meta-llama/Meta-Llama-3.1-8B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 48%
  300. 300 DeepSeek-R1-Distill-Qwen-14B fine-tuned on domain-specific datasets; bfloat16 48%
  301. 301 OpenBuddy openbuddy-deepseek-67b-v18.1-4k chat models (RLHF, DPO, IFT, ...); bfloat16 48%
  302. 302 HeshamHaroon Testv1 chat models (RLHF, DPO, IFT, ...); float16 48%
  303. 303 MaziyarPanahi Calme-4x7B-MoE-v0.2 base merges and moerges; bfloat16 47.9%
  304. 304 zhengr MixTAO-7Bx2-MoE-v8.1 base merges and moerges; bfloat16 47.9%
  305. 305 ibivibiv multimaster-7b-v6 ; ? 47.9%
  306. 306 inceptionai jais-family-2p7b-chat ; ? 47.9%
  307. 307 Kukedlc NeuralLLaMa-3-8b-DT-v0.1 base merges and moerges; float16 47.8%
  308. 308 princeton-nlp Llama-3-Instruct-8B-RDPO chat models (RLHF, DPO, IFT, ...); bfloat16 47.8%
  309. 309 OpenBuddy openbuddy-qwen2.5llamaify-7b-v23.1-200k chat models (RLHF, DPO, IFT, ...); bfloat16 47.8%
  310. 310 airev-ai Jais-Inception-7b-V0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 47.8%
  311. 311 jsfs11 MixtureofMerges-MoE-4x7b-v5 base merges and moerges; bfloat16 47.8%
  312. 312 kevin009 llamaRAGdrama chat models (RLHF, DPO, IFT, ...); float16 47.8%
  313. 313 Kukedlc NeuralSynthesis-7B-v0.1 base merges and moerges; bfloat16 47.8%
  314. 314 vicgalle Configurable-Hermes-3-Llama-3.1-8B chat models (RLHF, DPO, IFT, ...); float16 47.7%
  315. 315 AIDC-AI Marco-LLM-AR pretrained; bfloat16 47.7%
  316. 316 princeton-nlp Llama-3-Instruct-8B-SimPO chat models (RLHF, DPO, IFT, ...); bfloat16 47.7%
  317. 317 Kukedlc NeuralSynthesis-7B-v0.3 base merges and moerges; bfloat16 47.7%
  318. 318 saltlux luxia-21.4b-alignment-v1.0 chat models (RLHF, DPO, IFT, ...); bfloat16 47.7%
  319. 319 MaziyarPanahi Llama-3-8B-Instruct-v0.9 chat models (RLHF, DPO, IFT, ...); bfloat16 47.6%
  320. 320 nbeerbower gemma2-gutenberg-27B chat models (RLHF, DPO, IFT, ...); bfloat16 47.6%
  321. 321 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter3 chat models (RLHF, DPO, IFT, ...); bfloat16 47.5%
  322. 322 airev-ai Jais-Inception-7b-V0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 47.5%
  323. 323 allknowingroger MultiverseEx26-7B-slerp base merges and moerges; bfloat16 47.5%
  324. 324 Qwen/Qwen1.5-32B-Chat chat models (RLHF, DPO, IFT, ...); bfloat16 47.5%
  325. 325 liminerity M7-7b base merges and moerges; bfloat16 47.5%
  326. 326 princeton-nlp Llama-3-Base-8B-SFT-SLiC-HF chat models (RLHF, DPO, IFT, ...); bfloat16 47.5%
  327. 327 grimjim Llama-3.1-8B-Instruct-abliterated_via_adapter base merges and moerges; bfloat16 47.4%
  328. 328 inceptionai jais-adapted-13b-chat ; ? 47.4%
  329. 329 elmrc Juhaina-Instruct-v1 chat models (RLHF, DPO, IFT, ...); bfloat16 47.4%
  330. 330 utter-project EuroLLM-9B pretrained; bfloat16 47.4%
  331. 331 jsfs11 MixtureofMerges-MoE-4x7b-v4 base merges and moerges; bfloat16 47.4%
  332. 332 RubielLabarta LogoS-7Bx2-MoE-13B-v0.2 base merges and moerges; bfloat16 47.3%
  333. 333 inceptionai jais-family-30b-16k-chat ; ? 47.3%
  334. 334 ClusterlabAi GemmAr-7B-v1 chat models (RLHF, DPO, IFT, ...); float16 47.3%
  335. 335 princeton-nlp Llama-3-Base-8B-SFT-ORPO chat models (RLHF, DPO, IFT, ...); bfloat16 47.3%
  336. 336 NousResearch Hermes-3-Llama-3.1-8B chat models (RLHF, DPO, IFT, ...); bfloat16 47.3%
  337. 337 EpistemeAI Fireball-Alpaca-Llama-3.1-8B-Philos-DPO-200 chat models (RLHF, DPO, IFT, ...); float16 47.3%
  338. 338 meraGPT mera-mix-4x7B fine-tuned on domain-specific datasets; bfloat16 47.2%
  339. 339 mlabonne Daredevil-8B-abliterated fine-tuned on domain-specific datasets; bfloat16 47.2%
  340. 340 OpenBuddy openbuddy-qwen2.5llamaify-14b-v23.1-200k chat models (RLHF, DPO, IFT, ...); bfloat16 47.2%
  341. 341 sambanovasystems SambaLingo-Arabic-Base-70B ; ? 47.2%
  342. 342 dzakwan dzakwan-MoE-4x7b-Beta fine-tuned on domain-specific datasets; float16 47.1%
  343. 343 cognitivecomputations laserxtral fine-tuned on domain-specific datasets; bfloat16 47.1%
  344. 344 Orenguteng Llama-3.1-8B-Lexi-Uncensored fine-tuned on domain-specific datasets; bfloat16 47.1%
  345. 345 TIGER-Lab MAmmoTH2-8B-Plus fine-tuned on domain-specific datasets; bfloat16 47.1%
  346. 346 Qwen1.5-14B-Chat pretrained; bfloat16 47.1%
  347. 347 Kquant03 CognitiveFusion2-4x7B-BF16 base merges and moerges; bfloat16 47.1%
  348. 348 MaziyarPanahi Topxtral-4x7B-v0.1 fine-tuned on domain-specific datasets; bfloat16 47.1%
  349. 349 vicgalle Configurable-Llama-3.1-8B-Instruct chat models (RLHF, DPO, IFT, ...); float16 47.1%
  350. 350 thariqkhalid ara_mle_sft_300_8b ; bfloat16 47.1%
  351. 351 princeton-nlp Llama-3-Instruct-8B-RDPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 47.1%
  352. 352 migtissera Tess-3-Mistral-Nemo-12B fine-tuned on domain-specific datasets; bfloat16 47%
  353. 353 Ali-C137 Ge1H10M-0000 pretrained; bfloat16 47%
  354. 354 FreedomIntelligence AceGPT-13B-chat chat models (RLHF, DPO, IFT, ...); float16 47%
  355. 355 princeton-nlp Llama-3-Instruct-8B-DPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 47%
  356. 356 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter2 chat models (RLHF, DPO, IFT, ...); bfloat16 47%
  357. 357 Slim205 Barka-2b-it fine-tuned on domain-specific datasets; bfloat16 47%
  358. 358 EpistemeAI Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R2 fine-tuned on domain-specific datasets; float16 47%
  359. 359 princeton-nlp Llama-3-Instruct-8B-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 46.9%
  360. 360 bunnycore HyperLlama-3.1-8B base merges and moerges; bfloat16 46.9%
  361. 361 Kukedlc NeuralExperiment-7b-MagicCoder-v7.5 fine-tuned on domain-specific datasets; float16 46.9%
  362. 362 Orenguteng Llama-3.1-8B-Lexi-Uncensored-V2 fine-tuned on domain-specific datasets; bfloat16 46.9%
  363. 363 EpistemeAI Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R1-KTO-beta chat models (RLHF, DPO, IFT, ...); float16 46.9%
  364. 364 MTSAIR multi_verse_model fine-tuned on domain-specific datasets; bfloat16 46.9%
  365. 365 Kukedlc NeuralSynthesis-7b-v0.4-slerp base merges and moerges; bfloat16 46.9%
  366. 366 lightblue suzume-llama-3-8B-multilingual-orpo-borda-full chat models (RLHF, DPO, IFT, ...); bfloat16 46.9%
  367. 367 NousResearch Nous-Hermes-2-SOLAR-10.7B fine-tuned on domain-specific datasets; bfloat16 46.9%
  368. 368 automerger YamshadowExperiment28-7B base merges and moerges; bfloat16 46.8%
  369. 369 princeton-nlp Llama-3-Instruct-8B-RRHF-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 46.8%
  370. 370 EpistemeAI2 Fireball-Alpaca-Llama-3.1-8B-Instruct-KTO-beta chat models (RLHF, DPO, IFT, ...); float16 46.8%
  371. 371 NousResearch Hermes-2-Theta-Llama-3-8B chat models (RLHF, DPO, IFT, ...); bfloat16 46.8%
  372. 372 VAGOsolutions SauerkrautLM-Gemma-7b chat models (RLHF, DPO, IFT, ...); bfloat16 46.8%
  373. 373 princeton-nlp Llama-3-Instruct-8B-IPO chat models (RLHF, DPO, IFT, ...); bfloat16 46.8%
  374. 374 Kukedlc NeuralLLaMa-3-8b-ORPO-v0.3 chat models (RLHF, DPO, IFT, ...); float16 46.8%
  375. 375 EpistemeAI Fireball-Alpaca-Llama3.1.07-8B-Philos-Math-KTO-beta chat models (RLHF, DPO, IFT, ...); float16 46.7%
  376. 376 VAGOsolutions Llama-3-SauerkrautLM-8b-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 46.7%
  377. 377 yunconglong Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B chat models (RLHF, DPO, IFT, ...); bfloat16 46.7%
  378. 378 allknowingroger Ph3della5-14B base merges and moerges; float16 46.7%
  379. 379 mlabonne Beyonder-4x7B-v3 base merges and moerges; float16 46.7%
  380. 380 princeton-nlp Llama-3-Instruct-8B-SLiC-HF-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 46.7%
  381. 381 v000000 SwallowMaid-8B-L3-SPPO-abliterated ; ? 46.7%
  382. 382 princeton-nlp Llama-3-Instruct-8B-IPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 46.7%
  383. 383 DeepMount00 Llama-3-8b-Ita fine-tuned on domain-specific datasets; bfloat16 46.7%
  384. 384 princeton-nlp Llama-3-Instruct-8B-ORPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 46.7%
  385. 385 jpacifico Chocolatine-14B-Instruct-DPO-v1.2 chat models (RLHF, DPO, IFT, ...); float16 46.6%
  386. 386 mlabonne ChimeraLlama-3-8B-v3 base merges and moerges; float16 46.6%
  387. 387 Weyaxi Einstein-v7-Qwen2-7B chat models (RLHF, DPO, IFT, ...); bfloat16 46.6%
  388. 388 MaziyarPanahi calme-2.4-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 46.6%
  389. 389 MaziyarPanahi calme-2.6-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 46.6%
  390. 390 Qwen/Qwen2.5-Coder-7B-Instruct fine-tuned on domain-specific datasets; bfloat16 46.5%
  391. 391 OpenBuddy openbuddy-qwen1.5-32b-v21.1-32k chat models (RLHF, DPO, IFT, ...); bfloat16 46.5%
  392. 392 Qwen/Qwen2-MoE-57B-A14B ; bfloat16 46.5%
  393. 393 yunconglong MoE_13B_DPO chat models (RLHF, DPO, IFT, ...); bfloat16 46.5%
  394. 394 cognitivecomputations dolphin-2.9.2-qwen2-7b chat models (RLHF, DPO, IFT, ...); bfloat16 46.5%
  395. 395 MaziyarPanahi calme-2.2-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 46.5%
  396. 396 sophosympatheia Midnight-Miqu-70B-v1.5 base merges and moerges; float16 46.5%
  397. 397 MaziyarPanahi calme-2.1-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 46.5%
  398. 398 inceptionai jais-adapted-7b-chat ; ? 46.5%
  399. 399 Gemma 1.1 7B IT chat models (RLHF, DPO, IFT, ...); bfloat16 46.5%
  400. 400 elmrc JuhainaL2 chat models (RLHF, DPO, IFT, ...); bfloat16 46.4%
  401. 401 uygarkurt llama-3-merged-linear base merges and moerges; float16 46.4%
  402. 402 shadowml BeagSake-7B base merges and moerges; bfloat16 46.4%
  403. 403 jpacifico Chocolatine-14B-Instruct-4k-DPO chat models (RLHF, DPO, IFT, ...); float16 46.4%
  404. 404 mlabonne AlphaMonarch-7B chat models (RLHF, DPO, IFT, ...); float16 46.4%
  405. 405 OpenBuddy openbuddy-qwen1.5-32b-v21.2-32k chat models (RLHF, DPO, IFT, ...); bfloat16 46.3%
  406. 406 Qwen/Qwen2.5-3B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 46.3%
  407. 407 princeton-nlp Llama-3-Instruct-8B-KTO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 46.3%
  408. 408 elmrc JuhainaL chat models (RLHF, DPO, IFT, ...); bfloat16 46.3%
  409. 409 MaziyarPanahi Llama-3-8B-Instruct-v0.8 chat models (RLHF, DPO, IFT, ...); bfloat16 46.3%
  410. 410 phi-3-medium 14B chat models (RLHF, DPO, IFT, ...); bfloat16 46.3%
  411. 411 cstr llama3-8b-spaetzle-v37 base merges and moerges; bfloat16 46.2%
  412. 412 BAAI Infinity-Instruct-3M-0625-Llama3-8B chat models (RLHF, DPO, IFT, ...); bfloat16 46.2%
  413. 413 mlabonne NeuralMonarch-7B chat models (RLHF, DPO, IFT, ...); float16 46.1%
  414. 414 yunconglong DARE_TIES_13B base merges and moerges; bfloat16 46.1%
  415. 415 MaziyarPanahi calme-2.7-qwen2-7b fine-tuned on domain-specific datasets; bfloat16 46.1%
  416. 416 CultriX NeuralMona_MoE-4x7B base merges and moerges; bfloat16 46.1%
  417. 417 grimjim Llama-3-Instruct-8B-SimPO-SPPO-Iter3-merge base merges and moerges; bfloat16 46.1%
  418. 418 SeaLLMs SeaLLMs-v3-7B-Chat chat models (RLHF, DPO, IFT, ...); bfloat16 46.1%
  419. 419 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter1 chat models (RLHF, DPO, IFT, ...); bfloat16 46.1%
  420. 420 Qwen2.5-Coder-14B-Instruct fine-tuned on domain-specific datasets; bfloat16 46.1%
  421. 421 dnhkng RYS-Llama-3.1-8B-Instruct fine-tuned on domain-specific datasets; bfloat16 46%
  422. 422 Gemma 7B pretrained; bfloat16 46%
  423. 423 inceptionai jais-family-30b-16k ; ? 46%
  424. 424 Qwen2.5-Coder-14B pretrained; bfloat16 46%
  425. 425 macadeliccc Samantha-Qwen-2-7B fine-tuned on domain-specific datasets; float16 45.9%
  426. 426 freewheelin free-solar-evo-v0.13 fine-tuned on domain-specific datasets; float16 45.9%
  427. 427 leafspark Llama-3.1-8B-MultiReflection-Instruct chat models (RLHF, DPO, IFT, ...); float16 45.9%
  428. 428 mlabonne ChimeraLlama-3-8B-v2 base merges and moerges; float16 45.8%
  429. 429 Magpie-Align MagpieLM-8B-Chat-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 45.8%
  430. 430 maywell Qwen2-7B-Multilingual-RP fine-tuned on domain-specific datasets; bfloat16 45.8%
  431. 431 princeton-nlp Llama-3-Instruct-8B-KTO chat models (RLHF, DPO, IFT, ...); bfloat16 45.8%
  432. 432 inceptionai jais-family-30b-8k ; ? 45.8%
  433. 433 princeton-nlp Llama-3-Instruct-8B-CPO-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 45.8%
  434. 434 BAAI Infinity-Instruct-7M-Gen-Llama3_1-8B fine-tuned on domain-specific datasets; bfloat16 45.8%
  435. 435 wenbopan Faro-Yi-9B-DPO ; bfloat16 45.8%
  436. 436 freewheelin free-solar-evo-v0.11 fine-tuned on domain-specific datasets; float16 45.8%
  437. 437 Qwen1.5-32B pretrained; bfloat16 45.8%
  438. 438 Phi-3-medium-128k-instruct chat models (RLHF, DPO, IFT, ...); bfloat16 45.7%
  439. 439 lightblue suzume-llama-3-8B-multilingual-orpo-borda-top25 chat models (RLHF, DPO, IFT, ...); bfloat16 45.7%
  440. 440 EpistemeAI Fireball-Meta-Llama-3.1-8B-Instruct-Agent-0.003-128K-code fine-tuned on domain-specific datasets; float16 45.7%
  441. 441 gaverfraxz Meta-Llama-3.1-8B-Instruct-HalfAbliterated-TIES base merges and moerges; float16 45.6%
  442. 442 Weyaxi Bagel-Hermes-34B-Slerp base merges and moerges; bfloat16 45.6%
  443. 443 Kukedlc NeuralLLaMa-3-8b-ORPO-v0.4 chat models (RLHF, DPO, IFT, ...); float16 45.6%
  444. 444 HeshamHaroon llama-3-instruct-slerp-arabic ; bfloat16 45.6%
  445. 445 freewheelin free-solar-evo-v0.1 fine-tuned on domain-specific datasets; float16 45.5%
  446. 446 Ali-C137 Q2AW1M-0000 pretrained; bfloat16 45.5%
  447. 447 Ali-C137 Q2AW1M-1000 pretrained; bfloat16 45.5%
  448. 448 Ali-C137 Q2AW1M-1010 pretrained; bfloat16 45.5%
  449. 449 jondurbin bagel-dpo-34b-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 45.5%
  450. 450 Ali-C137 PG7BB pretrained; bfloat16 45.5%
  451. 451 Ali-C137 Q2H10M-0000 pretrained; bfloat16 45.5%
  452. 452 abacusai Smaug-Mixtral-v0.1 ; ? 45.5%
  453. 453 Danielbrdz Barcenas-Llama3-8b-ORPO chat models (RLHF, DPO, IFT, ...); float16 45.5%
  454. 454 Ali-C137 Q2AW1M-0010 pretrained; bfloat16 45.5%
  455. 455 Ali-C137 Q2AW1M-1100 pretrained; bfloat16 45.5%
  456. 456 abacusai bigstral-12b-32k base merges and moerges; float16 45.4%
  457. 457 chujiezheng Mistral7B-PairRM-SPPO-ExPO chat models (RLHF, DPO, IFT, ...); bfloat16 45.4%
  458. 458 cloudyu TomGrc_FusionNet_34Bx2_MoE_v0.1_DPO_f16 chat models (RLHF, DPO, IFT, ...); bfloat16 45.4%
  459. 459 mssqpi ALQLora2 ; ? 45.4%
  460. 460 Ali-C137 Q2AW1M-0100 pretrained; bfloat16 45.4%
  461. 461 ibivibiv llama-3-nectar-dpo-8B ; ? 45.3%
  462. 462 2A2I-R L500MT pretrained; bfloat16 45.3%
  463. 463 2A2I-R L100MT pretrained; bfloat16 45.3%
  464. 464 cloudyu Mixtral_34Bx2_MoE_60B base merges and moerges; bfloat16 45.3%
  465. 465 meta-llama/Meta-Llama-3-8B-Instruct fine-tuned on domain-specific datasets; float16 45.3%
  466. 466 Magpie-Align Llama-3-8B-Magpie-Align-SFT-v0.3 fine-tuned on domain-specific datasets; bfloat16 45.3%
  467. 467 2A2I-R L200MT pretrained; bfloat16 45.3%
  468. 468 2A2I-R L1000MT pretrained; bfloat16 45.2%
  469. 469 Magpie-Align Llama-3.1-8B-Magpie-Align-SFT-v0.1 fine-tuned on domain-specific datasets; bfloat16 45.2%
  470. 470 core42 jais-30b-v3 ; ? 45.2%
  471. 471 GritLM GritLM-8x7B fine-tuned on domain-specific datasets; bfloat16 45.2%
  472. 472 cloudyu Yi-34Bx2-MOE-200K base merges and moerges; bfloat16 45.1%
  473. 473 UCLA-AGI Mistral7B-PairRM-SPPO chat models (RLHF, DPO, IFT, ...); bfloat16 45.1%
  474. 474 UCLA-AGI Mistral7B-PairRM-SPPO-Iter2 chat models (RLHF, DPO, IFT, ...); bfloat16 45.1%
  475. 475 Magpie-Align MagpieLM-8B-SFT-v0.1 fine-tuned on domain-specific datasets; bfloat16 45.1%
  476. 476 princeton-nlp Llama-3-Base-8B-SFT-CPO chat models (RLHF, DPO, IFT, ...); bfloat16 45%
  477. 477 BAAI Infinity-Instruct-3M-0625-Qwen2-7B chat models (RLHF, DPO, IFT, ...); bfloat16 45%
  478. 478 abacusai Slerp-CM-mist-dpo chat models (RLHF, DPO, IFT, ...); float16 45%
  479. 479 princeton-nlp Mistral-7B-Instruct-RRHF chat models (RLHF, DPO, IFT, ...); bfloat16 44.9%
  480. 480 failspy Phi-3-medium-4k-instruct-abliterated-v3 fine-tuned on domain-specific datasets; bfloat16 44.9%
  481. 481 cloudyu Yi-34Bx2-MoE-60B base merges and moerges; bfloat16 44.9%
  482. 482 UCLA-AGI Mistral7B-PairRM-SPPO-Iter3 chat models (RLHF, DPO, IFT, ...); bfloat16 44.9%
  483. 483 UCLA-AGI Mistral7B-PairRM-SPPO-Iter1 chat models (RLHF, DPO, IFT, ...); bfloat16 44.8%
  484. 484 princeton-nlp Llama-3-Base-8B-SFT-RRHF chat models (RLHF, DPO, IFT, ...); bfloat16 44.8%
  485. 485 nvidia Llama3-ChatQA-1.5-8B chat models (RLHF, DPO, IFT, ...); float16 44.8%
  486. 486 princeton-nlp Llama-3-Instruct-8B-RRHF chat models (RLHF, DPO, IFT, ...); bfloat16 44.8%
  487. 487 princeton-nlp Mistral-7B-Instruct-SimPO chat models (RLHF, DPO, IFT, ...); bfloat16 44.7%
  488. 488 princeton-nlp Llama-3-Instruct-8B-CPO chat models (RLHF, DPO, IFT, ...); bfloat16 44.7%
  489. 489 failspy Codestral-22B-v0.1-abliterated-v3 fine-tuned on domain-specific datasets; bfloat16 44.6%
  490. 490 silma-ai SILMA-Kashif-2B-Instruct-v1.0 chat models (RLHF, DPO, IFT, ...); bfloat16 44.6%
  491. 491 cloudyu Yi-34Bx2-MoE-60B-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 44.6%
  492. 492 vicgalle Configurable-Llama-3-8B-v0.3 chat models (RLHF, DPO, IFT, ...); float16 44.6%
  493. 493 princeton-nlp Mistral-7B-Instruct-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 44.5%
  494. 494 cognitivecomputations dolphin-2.9.3-llama-3-8b fine-tuned on domain-specific datasets; bfloat16 44.5%
  495. 495 cognitivecomputations dolphin-2.9.4-llama3.1-8b fine-tuned on domain-specific datasets; bfloat16 44.5%
  496. 496 cognitivecomputations Llama-3-8B-Instruct-abliterated-v2 fine-tuned on domain-specific datasets; bfloat16 44.4%
  497. 497 cognitivecomputations dolphin-2.6-mixtral-8x7b fine-tuned on domain-specific datasets; bfloat16 44.3%
  498. 498 freewheelin free-llama3-dpo-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 44.3%
  499. 499 Ali-C137 Q2AW1M-1001 pretrained; bfloat16 44.3%
  500. 500 princeton-nlp Llama-3-Instruct-8B-SLiC-HF chat models (RLHF, DPO, IFT, ...); bfloat16 44.3%
  501. 501 01-ai/Yi-Coder-9B-Chat fine-tuned on domain-specific datasets; bfloat16 44.3%
  502. 502 Ali-C137 Q2AW1M-0001 pretrained; bfloat16 44.3%
  503. 503 Artples L-MChat-7b base merges and moerges; bfloat16 44.3%
  504. 504 cognitivecomputations dolphin-2.7-mixtral-8x7b fine-tuned on domain-specific datasets; bfloat16 44.2%
  505. 505 ConvexAI Luminex-34B-v0.2 ; bfloat16 44.2%
  506. 506 abacusai Smaug-34B-v0.1 chat models (RLHF, DPO, IFT, ...); bfloat16 44.1%
  507. 507 EpistemeAI2 Fireball-Llama-3.1-8B-Philos-Relection fine-tuned on domain-specific datasets; float16 44.1%
  508. 508 princeton-nlp Llama-3-Base-8B-SFT fine-tuned on domain-specific datasets; bfloat16 44.1%
  509. 509 mlabonne Meta-Llama-3.1-8B-Instruct-abliterated fine-tuned on domain-specific datasets; bfloat16 44.1%
  510. 510 OpenBuddy openbuddy-llama3.1-8b-v22.3-131k chat models (RLHF, DPO, IFT, ...); bfloat16 44.1%
  511. 511 princeton-nlp Llama-3-Instruct-8B-ORPO chat models (RLHF, DPO, IFT, ...); bfloat16 44.1%
  512. 512 fblgit UNA-SimpleSmaug-34b-v1beta base merges and moerges; bfloat16 44%
  513. 513 Omartificial-Intelligence-Space al-baka-llama3-8b-experimental fine-tuned on domain-specific datasets; float16 44%
  514. 514 lightblue suzume-llama-3-8B-multilingual-orpo-borda-half chat models (RLHF, DPO, IFT, ...); bfloat16 44%
  515. 515 princeton-nlp Mistral-7B-Base-SFT-KTO chat models (RLHF, DPO, IFT, ...); bfloat16 44%
  516. 516 vicgalle Configurable-Mistral-7B chat models (RLHF, DPO, IFT, ...); float16 44%
  517. 517 MaziyarPanahi Mistral-7B-Instruct-v0.3 fine-tuned on domain-specific datasets; bfloat16 43.9%
  518. 518 GritLM GritLM-7B-KTO chat models (RLHF, DPO, IFT, ...); bfloat16 43.9%
  519. 519 princeton-nlp Mistral-7B-Instruct-RDPO chat models (RLHF, DPO, IFT, ...); bfloat16 43.9%
  520. 520 Qwen2-7B pretrained; bfloat16 43.8%
  521. 521 ConvexAI Luminex-34B-v0.1 ; bfloat16 43.8%
  522. 522 mistralai/Mistral-7B-Instruct-v0.3 fine-tuned on domain-specific datasets; bfloat16 43.8%
  523. 523 princeton-nlp Mistral-7B-Instruct-CPO chat models (RLHF, DPO, IFT, ...); bfloat16 43.8%
  524. 524 tiiuae/falcon-180B-chat fine-tuned on domain-specific datasets; bfloat16 43.8%
  525. 525 inceptionai jais-family-1p3b-chat ; ? 43.8%
  526. 526 princeton-nlp Llama-3-8B-ProLong-64k-Base ; ? 43.7%
  527. 527 princeton-nlp Mistral-7B-Instruct-KTO chat models (RLHF, DPO, IFT, ...); bfloat16 43.7%
  528. 528 cognitivecomputations openchat-3.5-0106-laser fine-tuned on domain-specific datasets; bfloat16 43.7%
  529. 529 openchat openchat-3.5-0106 chat models (RLHF, DPO, IFT, ...); bfloat16 43.7%
  530. 530 Ali-C137 Q2AW1M-0111 pretrained; bfloat16 43.7%
  531. 531 arcee-ai Virtuoso-Lite chat models (RLHF, DPO, IFT, ...); bfloat16 43.6%
  532. 532 rombodawg rombos_Replete-Coder-Instruct-8b-Merged base merges and moerges; bfloat16 43.6%
  533. 533 Ali-C137 Q2AW1M-1111 pretrained; bfloat16 43.6%
  534. 534 Qwen2.5-3B pretrained; bfloat16 43.6%
  535. 535 cognitivecomputations dolphin-2.9.1-llama-3-8b fine-tuned on domain-specific datasets; bfloat16 43.5%
  536. 536 wenbopan Faro-Yi-34B-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 43.5%
  537. 537 Mistral-7B-Instruct-v0.2 chat models (RLHF, DPO, IFT, ...); bfloat16 43.5%
  538. 538 dfurman Llama-3-8B-Orpo-v0.1 chat models (RLHF, DPO, IFT, ...); float16 43.5%
  539. 539 failspy Meta-Llama-3-8B-Instruct-abliterated-v3 fine-tuned on domain-specific datasets; bfloat16 43.4%
  540. 540 ibivibiv llama3-8b-instruct-summary-combined ; ? 43.4%
  541. 541 Qwen2.5-Coder (7B) pretrained; bfloat16 43.3%
  542. 542 lightblue suzume-llama-3-8B-multilingual fine-tuned on domain-specific datasets; bfloat16 43.3%
  543. 543 jondurbin bagel-dpo-34b-v0.5 chat models (RLHF, DPO, IFT, ...); bfloat16 43.2%
  544. 544 01-ai/Yi-1.5-34B-Chat-16K fine-tuned on domain-specific datasets; bfloat16 43.1%
  545. 545 Qwen1.5-7B-Chat pretrained; bfloat16 43.1%
  546. 546 INSAIT-Institute BgGPT-7B-Instruct-v0.2 pretrained; bfloat16 43.1%
  547. 547 Ali-C137 L3H10M-0000 pretrained; bfloat16 43%
  548. 548 ValiantLabs Llama3.1-8B-ShiningValiant2 ; ? 43%
  549. 549 BAAI Infinity-Instruct-3M-0625-Yi-1.5-9B chat models (RLHF, DPO, IFT, ...); bfloat16 43%
  550. 550 Llama 3.1-8B pretrained; bfloat16 43%
  551. 551 princeton-nlp Mistral-7B-Instruct-IPO chat models (RLHF, DPO, IFT, ...); bfloat16 43%
  552. 552 meta-llama/Meta-Llama-3.1-8B pretrained; bfloat16 42.9%
  553. 553 vicgalle ConfigurableHermes-7B chat models (RLHF, DPO, IFT, ...); float16 42.9%
  554. 554 abacusai Llama-3-Smaug-8B fine-tuned on domain-specific datasets; bfloat16 42.9%
  555. 555 tokyotech-llm Llama-3-Swallow-8B-Instruct-v0.1 fine-tuned on domain-specific datasets; bfloat16 42.9%
  556. 556 anthracite-org magnum-v3-34b chat models (RLHF, DPO, IFT, ...); bfloat16 42.9%
  557. 557 migtissera Tess-v2.5-Phi-3-medium-128k-14B fine-tuned on domain-specific datasets; bfloat16 42.8%
  558. 558 meta-llama/Llama-3.2-3B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 42.7%
  559. 559 beowolx CodeNinja-1.0-OpenChat-7B fine-tuned on domain-specific datasets; bfloat16 42.6%
  560. 560 CombinHorizon YiSM-blossom5.1-34B-SLERP base merges and moerges; bfloat16 42.6%
  561. 561 princeton-nlp Mistral-7B-Instruct-ORPO chat models (RLHF, DPO, IFT, ...); bfloat16 42.6%
  562. 562 wenbopan Faro-Yi-9B fine-tuned on domain-specific datasets; bfloat16 42.5%
  563. 563 cognitivecomputations dolphin-2.9.1-yi-1.5-34b fine-tuned on domain-specific datasets; bfloat16 42.5%
  564. 564 SeaLLMs SeaLLM-7B-v2 fine-tuned on domain-specific datasets; bfloat16 42.5%
  565. 565 BAAI Infinity-Instruct-3M-0613-Mistral-7B fine-tuned on domain-specific datasets; bfloat16 42.4%
  566. 566 altomek YiSM-34B-0rn base merges and moerges; float16 42.4%
  567. 567 meta-llama/Meta-Llama-3-8B pretrained; bfloat16 42.4%
  568. 568 vicgalle Configurable-Janus-7B chat models (RLHF, DPO, IFT, ...); float16 42.3%
  569. 569 terrycraddock Reflection-Llama-3.1-8B chat models (RLHF, DPO, IFT, ...); float16 42.3%
  570. 570 FreedomIntelligence AceGPT-7B-chat chat models (RLHF, DPO, IFT, ...); float16 42.2%
  571. 571 01-ai/Yi-Coder-9B pretrained; bfloat16 42.2%
  572. 572 vicgalle Configurable-Yi-1.5-9B-Chat chat models (RLHF, DPO, IFT, ...); float16 42.2%
  573. 573 abacusai Liberated-Qwen1.5-14B fine-tuned on domain-specific datasets; float16 42.2%
  574. 574 upstage SOLAR-10.7B-v1.0 pretrained; float16 42.2%
  575. 575 google/gemma-7b-it fine-tuned on domain-specific datasets; bfloat16 42.1%
  576. 576 Falcon3-10B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 42.1%
  577. 577 princeton-nlp Mistral-7B-Instruct-SLiC-HF chat models (RLHF, DPO, IFT, ...); bfloat16 42.1%
  578. 578 01-ai/Yi-1.5-34B-Chat chat models (RLHF, DPO, IFT, ...); bfloat16 42.1%
  579. 579 01-ai/Yi-1.5-9B-Chat-16K fine-tuned on domain-specific datasets; bfloat16 42%
  580. 580 cognitivecomputations dolphin-2.9.3-mistral-7B-32k fine-tuned on domain-specific datasets; bfloat16 42%
  581. 581 Qwen1.5-14B pretrained; bfloat16 41.9%
  582. 582 Yi-1.5-9B-Chat chat models (RLHF, DPO, IFT, ...); bfloat16 41.9%
  583. 583 oobabooga CodeBooga-34B-v0.1 fine-tuned on domain-specific datasets; float16 41.9%
  584. 584 brucethemoose Yi-34B-200K-DARE-merge-v7 base merges and moerges; bfloat16 41.8%
  585. 585 unsloth llama-3-8b-bnb-4bit ; GPTQ 41.7%
  586. 586 oddadmix Arabic-Qwen2.5-3B fine-tuned on domain-specific datasets; bfloat16 41.7%
  587. 587 wenbopan Faro-Yi-34B chat models (RLHF, DPO, IFT, ...); bfloat16 41.7%
  588. 588 chujiezheng Smaug-34B-v0.1-ExPO fine-tuned on domain-specific datasets; bfloat16 41.7%
  589. 589 meta-llama/Llama-2-70b-hf pretrained; float16 41.6%
  590. 590 Weyaxi Nous-Hermes-2-SUS-Chat-34B-Slerp base merges and moerges; bfloat16 41.5%
  591. 591 royallab ZephRP-m7b base merges and moerges; bfloat16 41.5%
  592. 592 byroneverson Yi-1.5-9B-Chat-16K-abliterated fine-tuned on domain-specific datasets; bfloat16 41.4%
  593. 593 byroneverson Yi-1.5-9B-Chat-abliterated fine-tuned on domain-specific datasets; bfloat16 41.4%
  594. 594 Ali-C137 Q15H10M-0000 pretrained; bfloat16 41.4%
  595. 595 cognitivecomputations dolphin-2.9.3-mistral-nemo-12b fine-tuned on domain-specific datasets; bfloat16 41.4%
  596. 596 sambanovasystems SambaLingo-Arabic-Chat chat models (RLHF, DPO, IFT, ...); bfloat16 41.3%
  597. 597 Falcon3-7B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 41.3%
  598. 598 cognitivecomputations dolphin-2.9.1-yi-1.5-9b fine-tuned on domain-specific datasets; bfloat16 41.2%
  599. 599 Omartificial-Intelligence-Space Arabic-llama3.1-16bit-FT ; bfloat16 41.1%
  600. 600 speakleash Bielik-11B-v2.3-Instruct base merges and moerges; float16 41.1%
  601. 601 rombodawg Everyone-Coder-4x7b-Base ; bfloat16 41.1%
  602. 602 OpenBuddy openbuddy-zero-56b-v21.2-32k chat models (RLHF, DPO, IFT, ...); bfloat16 41%
  603. 603 Azure99 blossom-v5.1-34b chat models (RLHF, DPO, IFT, ...); bfloat16 41%
  604. 604 Falcon-180B pretrained; bfloat16 41%
  605. 605 FreedomIntelligence AceGPT-13B pretrained; float16 41%
  606. 606 qnguyen3 Master-Yi-9B chat models (RLHF, DPO, IFT, ...); bfloat16 41%
  607. 607 FreedomIntelligence AceGPT-v1.5-13B ; ? 41%
  608. 608 oddadmix Arabic-Qwen2.5-3B-v2.1 fine-tuned on domain-specific datasets; bfloat16 41%
  609. 609 uukuguy speechless-code-mistral-7b-v1.0 fine-tuned on domain-specific datasets; float16 40.9%
  610. 610 ghost-x ghost-8b-beta-1608 chat models (RLHF, DPO, IFT, ...); bfloat16 40.8%
  611. 611 inceptionai jais-adapted-13b ; ? 40.8%
  612. 612 618AI dictalm2-it-qa-fine-tune ; bfloat16 40.8%
  613. 613 deepseek-ai/DeepSeek-R1-Distill-Llama-8B fine-tuned on domain-specific datasets; bfloat16 40.7%
  614. 614 Qwen1.5-7B pretrained; bfloat16 40.7%
  615. 615 Qwen/Qwen2.5-Math-7B pretrained; bfloat16 40.6%
  616. 616 rombodawg rombos_Replete-Coder-Llama3-8B fine-tuned on domain-specific datasets; bfloat16 40.6%
  617. 617 uukuguy speechless-zephyr-code-functionary-7b fine-tuned on domain-specific datasets; bfloat16 40.4%
  618. 618 OpenBuddy openbuddy-yi1.5-34b-v21.3-32k chat models (RLHF, DPO, IFT, ...); bfloat16 40.4%
  619. 619 princeton-nlp Mistral-7B-Base-SFT-SimPO chat models (RLHF, DPO, IFT, ...); bfloat16 40.4%
  620. 620 aws-prototyping MegaBeam-Mistral-7B-512k fine-tuned on domain-specific datasets; bfloat16 40.4%
  621. 621 cognitivecomputations dolphin-2.9.2-Phi-3-Medium chat models (RLHF, DPO, IFT, ...); bfloat16 40.4%
  622. 622 SakanaAI DiscoPOP-zephyr-7b-gemma chat models (RLHF, DPO, IFT, ...); bfloat16 40.4%
  623. 623 GritLM GritLM-7B fine-tuned on domain-specific datasets; bfloat16 40.3%
  624. 624 Qwen/Qwen2.5-1.5B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 40.3%
  625. 625 Qwen2-1.5B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 40.3%
  626. 626 MohamedRashad Arabic-Orpo-Llama-3-8B-Instruct fine-tuned on domain-specific datasets; float16 40.3%
  627. 627 speakleash Bielik-11B-v2 pretrained; bfloat16 40.3%
  628. 628 kaist-ai janus-dpo-7b chat models (RLHF, DPO, IFT, ...); bfloat16 40.3%
  629. 629 princeton-nlp Mistral-7B-Base-SFT-RDPO chat models (RLHF, DPO, IFT, ...); bfloat16 40.2%
  630. 630 cognitivecomputations dolphin-2.9.2-Phi-3-Medium-abliterated fine-tuned on domain-specific datasets; bfloat16 40.2%
  631. 631 TIGER-Lab MAmmoTH2-7B-Plus fine-tuned on domain-specific datasets; bfloat16 40.1%
  632. 632 kaist-ai janus-7b fine-tuned on domain-specific datasets; bfloat16 40.1%
  633. 633 cognitivecomputations dolphin-2.9.3-Yi-1.5-34B-32k chat models (RLHF, DPO, IFT, ...); bfloat16 40.1%
  634. 634 kaist-ai mistral-orpo-capybara-7k chat models (RLHF, DPO, IFT, ...); bfloat16 40%
  635. 635 princeton-nlp Mistral-7B-Base-SFT-SLiC-HF chat models (RLHF, DPO, IFT, ...); bfloat16 40%
  636. 636 uukuguy zephyr-7b-alpha-dare-0.85 fine-tuned on domain-specific datasets; float16 40%
  637. 637 cognitivecomputations dolphin-2.9.1-mixtral-1x22b fine-tuned on domain-specific datasets; bfloat16 39.9%
  638. 638 oddadmix Arabic-Qwen2.5-3B-v2 fine-tuned on domain-specific datasets; bfloat16 39.8%
  639. 639 CausalLM 34b-beta fine-tuned on domain-specific datasets; bfloat16 39.8%
  640. 640 google/recurrentgemma-2b-it chat models (RLHF, DPO, IFT, ...); bfloat16 39.8%
  641. 641 amazon MegaBeam-Mistral-7B-300k fine-tuned on domain-specific datasets; bfloat16 39.7%
  642. 642 core42 jais-30b-v1 ; ? 39.6%
  643. 643 Qwen2.5-1.5B pretrained; bfloat16 39.6%
  644. 644 mlabonne FineLlama-3.1-8B fine-tuned on domain-specific datasets; bfloat16 39.6%
  645. 645 01-ai/Yi-1.5-9B pretrained; bfloat16 39.5%
  646. 646 NLPark AnFeng_v3.1-Avocet chat models (RLHF, DPO, IFT, ...); float16 39.5%
  647. 647 01-ai/Yi-1.5-34B-32K pretrained; bfloat16 39.5%
  648. 648 01-ai/Yi-1.5-9B-32K pretrained; bfloat16 39.5%
  649. 649 01-ai/Yi-9B-200K pretrained; bfloat16 39.5%
  650. 650 tiiuae/Falcon3-10B-Base pretrained; bfloat16 39.4%
  651. 651 maywell Synatra-7B-v0.3-RP fine-tuned on domain-specific datasets; float16 39.4%
  652. 652 Qwen/Qwen1.5-MoE-A2.7B pretrained; bfloat16 39.4%
  653. 653 Qwen/Qwen1.5-MoE-A2.7B-Chat pretrained; bfloat16 39.4%
  654. 654 princeton-nlp Mistral-7B-Base-SFT-RRHF chat models (RLHF, DPO, IFT, ...); bfloat16 39.3%
  655. 655 Ali-C137 M03H10M-0000 pretrained; bfloat16 39.2%
  656. 656 Sakalti Saka-1.5B base merges and moerges; float16 39.1%
  657. 657 OpenBuddy openbuddy-yi1.5-34b-v21.6-32k-fp16 chat models (RLHF, DPO, IFT, ...); float16 39.1%
  658. 658 princeton-nlp Mistral-7B-Base-SFT-CPO chat models (RLHF, DPO, IFT, ...); bfloat16 39%
  659. 659 dicta-il dictalm2.0 pretrained; bfloat16 39%
  660. 660 princeton-nlp Mistral-7B-Base-SFT-IPO chat models (RLHF, DPO, IFT, ...); bfloat16 39%
  661. 661 Yi-1.5-34B pretrained; bfloat16 38.9%
  662. 662 NotAiLOL Yi-1.5-dolphin-9B fine-tuned on domain-specific datasets; bfloat16 38.9%
  663. 663 DeepSeek-R1-Distill-Qwen-7B fine-tuned on domain-specific datasets; bfloat16 38.8%
  664. 664 Syed-Hasan-8503 Phi-3-mini-4K-instruct-cpo-simpo chat models (RLHF, DPO, IFT, ...); bfloat16 38.8%
  665. 665 Mistral 7B v0.3 pretrained; bfloat16 38.7%
  666. 666 MaziyarPanahi Mistral-7B-v0.3 pretrained; bfloat16 38.7%
  667. 667 inceptionai jais-adapted-7b ; ? 38.6%
  668. 668 dicta-il dictalm2.0-instruct chat models (RLHF, DPO, IFT, ...); bfloat16 38.6%
  669. 669 princeton-nlp Mistral-7B-Base-SFT-DPO chat models (RLHF, DPO, IFT, ...); bfloat16 38.6%
  670. 670 sambanovasystems SambaLingo-Arabic-Base ; ? 38.5%
  671. 671 SenseLLM ReflectionCoder-CL-34B chat models (RLHF, DPO, IFT, ...); bfloat16 38.3%
  672. 672 inceptionai jais-family-13b ; ? 38.3%
  673. 673 kaist-ai janus-orpo-7b chat models (RLHF, DPO, IFT, ...); bfloat16 38.3%
  674. 674 BAAI OPI-Llama-3.1-8B-Instruct fine-tuned on domain-specific datasets; bfloat16 38.2%
  675. 675 migtissera Tess-M-v1.3 fine-tuned on domain-specific datasets; float16 38.2%
  676. 676 Gemma 1.1 2B IT chat models (RLHF, DPO, IFT, ...); bfloat16 38.1%
  677. 677 sajaw llama-2-7b-RandomGPT-5K-ar ; bfloat16 38%
  678. 678 FreedomIntelligence AceGPT-7B pretrained; float16 38%
  679. 679 abacusai bigyi-15b base merges and moerges; float16 38%
  680. 680 tiiuae/Falcon3-7B-Base pretrained; bfloat16 38%
  681. 681 core42 jais-13b ; ? 37.9%
  682. 682 haoranxu ALMA-13B-Pretrain ; ? 37.9%
  683. 683 Qwen2-1.5B pretrained; bfloat16 37.9%
  684. 684 google/gemma-2b-it fine-tuned on domain-specific datasets; bfloat16 37.8%
  685. 685 Qwen/Qwen1.5-4B pretrained; bfloat16 37.7%
  686. 686 meta-llama/Llama-2-13b-hf pretrained; float16 37.6%
  687. 687 gpt2 ; GPTQ 37.6%
  688. 688 inceptionai jais-family-6p7b ; ? 37.6%
  689. 689 Qwen/Qwen1.5-1.8B-Chat pretrained; bfloat16 37.5%
  690. 690 Qwen/Qwen2.5-Math-7B-Instruct fine-tuned on domain-specific datasets; bfloat16 37.5%
  691. 691 google/recurrentgemma-2b ; ? 37.4%
  692. 692 Qwen1.5-4B-Chat pretrained; bfloat16 37.3%
  693. 693 Ali-C137 L2H10M-0000 pretrained; bfloat16 37.3%
  694. 694 meta-llama/Llama-2-7b-hf pretrained; float16 37.2%
  695. 695 tiiuae/Falcon3-1B-Base pretrained; bfloat16 37%
  696. 696 SenseLLM ReflectionCoder-DS-33B chat models (RLHF, DPO, IFT, ...); bfloat16 37%
  697. 697 inceptionai jais-family-590m-chat ; ? 36.8%
  698. 698 Qwen/Qwen1.5-1.8B pretrained; bfloat16 36.7%
  699. 699 haoranxu ALMA-13B-R chat models (RLHF, DPO, IFT, ...); float16 36.6%
  700. 700 Qwen2-0.5B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 36.5%
  701. 701 byroneverson Yi-1.5-34B-Chat-abliterated fine-tuned on domain-specific datasets; bfloat16 36.5%
  702. 702 NTQAI Nxcode-CQ-7B-orpo chat models (RLHF, DPO, IFT, ...); bfloat16 36.5%
  703. 703 Qwen2.5-0.5B pretrained; bfloat16 36.4%
  704. 704 inceptionai jais-family-2p7b ; ? 36.3%
  705. 705 Qwen/Qwen1.5-0.5B-Chat pretrained; bfloat16 36.2%
  706. 706 Llama 3.2 1B pretrained; bfloat16 36.2%
  707. 707 haoranxu ALMA-7B-R ; float16 36.1%
  708. 708 Falcon3-3B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 36.1%
  709. 709 Falcon3-1B-Instruct chat models (RLHF, DPO, IFT, ...); bfloat16 36%
  710. 710 inceptionai jais-family-1p3b ; ? 36%
  711. 711 Gemma 2B pretrained; bfloat16 36%
  712. 712 riotu-lab ArabianGPT-0.3B-QA ; ? 36%
  713. 713 inceptionai jais-family-590m ; ? 35.9%
  714. 714 bigscience bloom-7b1 pretrained; float16 35.9%
  715. 715 tiiuae/falcon-11B pretrained; bfloat16 35.7%
  716. 716 Qwen/Qwen1.5-0.5B pretrained; bfloat16 35.7%
  717. 717 riotu-lab ArabianGPT-01B ; ? 35.7%
  718. 718 riotu-lab ArabianGPT-03B ; ? 35.5%
  719. 719 tiiuae/falcon-7b-instruct fine-tuned on domain-specific datasets; bfloat16 35.5%
  720. 720 Falcon-40B pretrained; bfloat16 35.5%
  721. 721 riotu-lab ArabianGPT-08B-V2 ; ? 35.4%
  722. 722 kevinpro Hydra-LLaMA3-8B-v0.1 ; ? 35.4%
  723. 723 Qwen2-0.5B pretrained; bfloat16 35.3%
  724. 724 Artples L-MChat-Small base merges and moerges; bfloat16 35.2%
  725. 725 tiiuae/Falcon3-3B-Base pretrained; bfloat16 35.2%
  726. 726 kevinpro Hydra-LLaMA3-8B-v0.2 ; ? 35.1%
  727. 727 Falcon-7B pretrained; bfloat16 35%
  728. 728 Ali-C137 F1H10M-0000 pretrained; bfloat16 35%
  729. 729 paloalma ECE-TW3-JRGL-V3 base merges and moerges; bfloat16 35%
  730. 730 paloalma ECE-TW3-JRGL-V4 base merges and moerges; bfloat16 34.8%
  731. 731 riotu-lab ArabianGPT-1.5B ; ? 34.7%
  732. 732 tiiuae/falcon-40b-instruct fine-tuned on domain-specific datasets; bfloat16 34.6%
  733. 733 kaist-ai janus-rm-7b fine-tuned on domain-specific datasets; bfloat16 34.5%
  734. 734 MBZUAI MobiLlama-1B ; ? 34.4%

Source: Official Open Arabic LLM Leaderboard v1 leaderboard

#EntryScoreSettingDateSource
1 silma-ai SILMA-9B-Instruct-v1.0 70.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
2 MaziyarPanahi calme-2.1-qwen2.5-72b 69.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
3 Sakalti ultiima-72B 69.1% fine-tuned on domain-specific datasets; float16 Official leaderboard
4 Replete-AI Replete-LLM-V2.5-Qwen-72b 69.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
5 MaziyarPanahi calme-2.2-qwen2.5-72b 69.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
6 rombodawg Rombos-LLM-V2.5-Qwen-72b 69.1% base merges and moerges; bfloat16 Official leaderboard
7 Qwen2.5-72B 69.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Sep 2024 Official leaderboard
8 zetasepic Qwen2.5-72B-Instruct-abliterated 68.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
9 paloalma TW3-JRGL-v2 67.1% base merges and moerges; bfloat16 Official leaderboard
10 davidkim205 Rhea-72b-v0.5 67% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
11 AXCXEPT EZO-Qwen2.5-72B-Instruct 66.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
12 freewheelin free-evo-qwen72b-v0.8-re 65.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
13 MaziyarPanahi calme-2.2-qwen2-72b 65.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
14 dnhkng RYS-XLarge 65.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
15 arcee-ai Arcee-Nova 65.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
16 dnhkng RYS-XLarge-base 65% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
17 MTSAIR MultiVerse_70B 65% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
18 anthracite-org magnum-v1-72b 64.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
19 Undi95 MG-FinalMix-72B 64.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
20 MaziyarPanahi calme-2.1-qwen2-72b 64.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
21 dfurman Qwen2-72B-Orpo-v0.1 64.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
22 MaziyarPanahi Qwen2-72B-Instruct-v0.1 64.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
23 alpindale magnum-72b-v1 64.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
24 abacusai Dracarys-72B-Instruct 64.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
25 silma-ai SILMA-9B-Instruct-v0.1 64.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
26 Nexusflow Athene-70B 64.4% ; bfloat16 Official leaderboard
27 mssqpi G3 64.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
28 silma-ai SILMA-9B-Instruct-v0.1.0 64.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
29 silma-ai SILMA-9B-Instruct-v0.1.1 64.3% pretrained; bfloat16 Official leaderboard
30 tanliboy lambda-qwen2.5-32b-dpo-test 64% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
31 Sakalti ultiima-125B 64% fine-tuned on domain-specific datasets; float16 Official leaderboard
32 anthracite-org magnum-v2-72b 63.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
33 Qwen2-72B 63.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Jun 2024 Official leaderboard
34 paloalma Le_Triomphant-ECE-TW3 63.5% base merges and moerges; bfloat16 Official leaderboard
35 abacusai Smaug-Qwen2-72B-Instruct 63.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
36 pankajmathur orca_mini_v7_72b 63.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
37 Qwen/Qwen2.5-32B-Instruct 63.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
38 paulml ECE-ILAB-Q1 63% base merges and moerges; bfloat16 Official leaderboard
39 Sakalti ultiima-32B 63% base merges and moerges; float16 Official leaderboard
40 Replete-AI Replete-LLM-V2.5-Qwen-32b 62.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
41 VAGOsolutions Llama-3-SauerkrautLM-70b-Instruct 62.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
42 Sakalti ultiima-78B 62.7% fine-tuned on domain-specific datasets; float16 Official leaderboard
43 dfurman CalmeRys-78B-Orpo-v0.1 62.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
44 sometimesanotion Qwen2.5-14B-Vimarckoso-v3 62.4% base merges and moerges; bfloat16 Official leaderboard
45 sometimesanotion Lamarck-14B-v0.6 62.3% base merges and moerges; bfloat16 Official leaderboard
46 Sakalti ultiima-78B-V2 62.3% base merges and moerges; float16 Official leaderboard
47 hotmailuser QwenSlerp2-14B 62.3% base merges and moerges; bfloat16 Official leaderboard
48 AXCXEPT EZO-Qwen2.5-32B-Instruct 62.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
49 Sakalti SakaMoe-3x14B-Instruct 62.1% fine-tuned on domain-specific datasets; float16 Official leaderboard
50 tanliboy lambda-qwen2.5-14b-dpo-test 62% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
51 huihui-ai Qwen2.5-32B-Instruct-abliterated 61.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
52 sometimesanotion Lamarck-14B-v0.7 61.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
53 Slim205 Barka-9b-it-v02 61.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
54 airev-ai Amal-70b-v2 61.7% fine-tuned on domain-specific datasets; float16 Official leaderboard
55 airev-ai Amal-70b-v1.1 61.6% ; float16 Official leaderboard
56 migtissera Tess-v2.5.2-Qwen2-72B 61.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
57 airev-ai Amal-75b-v3 61.5% ; float16 Official leaderboard
58 byroneverson gemma-2-27b-it-abliterated 61.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
59 MaziyarPanahi calme-2.4-llama3-70b 61.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
60 Lambent qwen2.5-reinstruct-alternate-lumen-14B 61.3% base merges and moerges; bfloat16 Official leaderboard
61 v000000 Qwen2.5-Lumen-14B 61.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
62 MaziyarPanahi calme-2.2-llama3-70b 61.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
63 rombodawg Rombos-LLM-V2.6-Qwen-14b 61.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
64 Replete-AI Replete-LLM-V2.5-Qwen-14b 61.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
65 Sakalti ultiima-14B 61.1% base merges and moerges; float16 Official leaderboard
66 v000000 Qwen2.5-14B-Gutenberg-1e-Delta 61.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
67 airev-ai qwen-14b-v1 61.1% ; float16 Official leaderboard
68 Qwen/Qwen2.5-14B-Instruct 61.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
69 v000000 Qwen2.5-14B-Gutenberg-Instruct-Slerpeno 61% base merges and moerges; bfloat16 Official leaderboard
70 VAGOsolutions Llama-3.1-SauerkrautLM-70b-Instruct 60.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
71 FiditeNemini Unhinged-Qwen2-72B 60.9% base merges and moerges; bfloat16 Official leaderboard
72 yellowtown 7B-v0.2 60.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
73 MaziyarPanahi calme-2.3-llama3-70b 60.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
74 NousResearch Hermes-3-Llama-3.1-70B 60.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
75 google/gemma-2-27b-it 60.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
76 elmrc JuhainaG2 60.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
77 MaziyarPanahi Llama-3-70B-Instruct-DPO-v0.2 60.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
78 mmnga Llama-3-70B-japanese-suzume-vector-v0.1 60% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
79 airev-ai Jais-Inception-70b-V1.2.2 59.9% ; float16 Official leaderboard
80 meta-llama/Meta-Llama-3-70B-Instruct 59.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
81 failspy llama-3-70B-Instruct-abliterated 59.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
82 dnhkng RYS-Gemma-2-27b-it 59.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
83 airev-ai qwen-14b-v2 59.5% ; float16 Official leaderboard
84 airev-ai Jais-Inception-70b-V1.1 59.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
85 recoilme recoilme-gemma-2-9B-v0.4 59.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
86 cognitivecomputations dolphin-2.9.2-qwen2-72b 59.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
87 migtissera Tess-v2.5-Qwen2-72B 59% fine-tuned on domain-specific datasets; float16 Official leaderboard
88 Qwen2.5-14B-Instruct-1M 58.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Jan 2025 Official leaderboard
89 abacusai Smaug-72B-v0.1 58.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
90 elmrc juhainaG 58.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
91 princeton-nlp gemma-2-9b-it-SimPO 58.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
92 elmresearchcenter juhaina 58.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
93 gbueno86 Meta-LLama-3-Cat-Smaug-LLama-70b 58.4% base merges and moerges; bfloat16 Official leaderboard
94 Nohobby Q2.5-Qwetiapin-32B 58.2% base merges and moerges; bfloat16 Official leaderboard
95 anthracite-org magnum-v3-27b-kto 58.2% ; ? Official leaderboard
96 nvidia Llama-3.1-Nemotron-70B-Instruct-HF 58% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
97 Daemontatox Mawared_GS 58% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
98 airev-ai Amal-70b-v5 58% ; float16 Official leaderboard
99 arcee-ai Virtuoso-Small 57.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
100 maldv Qwenstein2.5-32B-Instruct 57.8% base merges and moerges; bfloat16 Official leaderboard
101 abhishek autotrain-llama3-70b-orpo-v1 57.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
102 gbueno86 Meta-LLama-3-Cat-A-LLama-70b 57.6% base merges and moerges; bfloat16 Official leaderboard
103 abhishek autotrain-llama3-70b-orpo-v2 57.5% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
104 VAGOsolutions SauerkrautLM-Nemo-12b-Instruct 57.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
105 UCLA-AGI Gemma-2-9B-It-SPPO-Iter3 57.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
106 MaziyarPanahi calme-2.3-llama3.1-70b 57.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
107 UCLA-AGI Gemma-2-9B-It-SPPO-Iter1 57% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
108 airev-ai emirati-14b-v2 57% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
109 google/gemma-2-9b-it 57% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
110 Ashmal MBZUAI-ORYX-new 57% ; float16 Official leaderboard
111 airev-ai Jais-Inception-70b-V1.2 57% ; float16 Official leaderboard
112 recoilme Gemma-2-Ataraxy-Gemmasutra-9B-slerp 57% base merges and moerges; bfloat16 Official leaderboard
113 UCLA-AGI Gemma-2-9B-It-SPPO-Iter2 57% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
114 MaziyarPanahi calme-2.2-llama3.1-70b 56.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
115 cognitivecomputations dolphin-2.9.1-llama-3-70b 56.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
116 rombodawg Rombos-LLM-V2.6-Nemotron-70b 56.9% base merges and moerges; bfloat16 Official leaderboard
117 inceptionai jais-adapted-70b-chat 56.8% ; ? Official leaderboard
118 elmrc JuhainaG9B 56.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
119 failspy Llama-3-70B-Instruct-abliterated-v3 56.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
120 shenzhi-wang Llama3-70B-Chinese-Chat 56.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
121 mssqpi Waie1 56.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
122 TareksLab Thalassic-Alpha-LLaMa-70B 56.8% base merges and moerges; bfloat16 Official leaderboard
123 airev-ai Amal-70b-v4 56.8% ; float16 Official leaderboard
124 hotmailuser Deeepseek-QwenSlerp4-32B 56.6% base merges and moerges; bfloat16 Official leaderboard
125 sambanovasystems SambaLingo-Arabic-Chat-70B 56.6% ; ? Official leaderboard
126 recoilme recoilme-gemma-2-9B-v0.2 56.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
127 Steelskull L3-MS-Astoria-70b 56.4% base merges and moerges; bfloat16 Official leaderboard
128 MaziyarPanahi calme-2.3-qwen2-72b 56.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
129 CohereForAI/c4ai-command-r-plus 56.4% fine-tuned on domain-specific datasets; float16 Official leaderboard
130 meta-llama/Meta-Llama-3.1-70B-Instruct 56.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
131 tenyx Llama3-TenyxChat-70B 56.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
132 BAAI Gemma2-9B-IT-Simpo-Infinity-Preference 55.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
133 inceptionai jais-family-30b-8k-chat 55.9% ; ? Official leaderboard
134 Ashmal MBZUAI-oryx 55.8% fine-tuned on domain-specific datasets; float16 Official leaderboard
135 migtissera Llama-3-70B-Synthia-v3.5 55.7% fine-tuned on domain-specific datasets; float16 Official leaderboard
136 CohereForAI/c4ai-command-r-plus-08-2024 55.7% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
137 princeton-nlp gemma-2-9b-it-DPO 55.7% ; bfloat16 Official leaderboard
138 recoilme recoilme-gemma-2-9B-v0.3 55.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
139 Triangle104 DeepSky-T100 55.3% base merges and moerges; float16 Official leaderboard
140 airev-ai emirati-14b-v3 55.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
141 Gemma 2 27B 55.2% ; ? Jun 2024 Official leaderboard
142 huihui-ai Qwen2.5-7B-Instruct-abliterated-v2 55% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
143 Delta-Vector Odin-9B 55% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
144 FreedomIntelligence AceGPT-v2-70B 55% pretrained; float16 Official leaderboard
145 prithivMLmods Qwen2.5-32B-DeepSeek-R1-Instruct 54.9% base merges and moerges; bfloat16 Official leaderboard
146 Joseph717171 Llama-3.1-SuperNova-8B-Lite_TIES_with_Base 54.9% base merges and moerges; bfloat16 Official leaderboard
147 Ba2han QwQenSeek-coder 54.8% base merges and moerges; bfloat16 Official leaderboard
148 FreedomIntelligence AceGPT-v2-32B-Chat 54.6% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
149 core42 jais-30b-chat-v3 54.5% ; ? Official leaderboard
150 Orion-zhen Qwen2.5-7B-Instruct-Uncensored 54.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
151 anthracite-org magnum-v4-9b 54.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
152 Command R (03-2024) 54.4% fine-tuned on domain-specific datasets; float16 Mar 2024 Official leaderboard
153 Isaak-Carter Josiefied-Qwen2.5-7B-Instruct-abliterated 54.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
154 BAAI Infinity-Instruct-7M-Gen-Llama3_1-70B 54.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
155 Qwen/Qwen2.5-7B-Instruct 54.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
156 Qwen2.5-32B 54.2% pretrained; bfloat16 Sep 2024 Official leaderboard
157 ai21labs AI21-Jamba-1.5-Mini 54.2% pretrained; bfloat16 Official leaderboard
158 anthracite-org magnum-v3-9b-customgemma2 54.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
159 Replete-AI Replete-LLM-V2.5-Qwen-7b 54.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
160 TheDrummer Gemmasutra-9B-v1 54.1% base merges and moerges; bfloat16 Official leaderboard
161 migtissera Tess-2.0-Llama-3-70B 54.1% fine-tuned on domain-specific datasets; float16 Official leaderboard
162 jondurbin airoboros-dpo-70b-3.3 54% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
163 failspy Meta-Llama-3-70B-Instruct-abliterated-v3.5 54% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
164 migtissera Tess-2.0-Llama-3-70B-v0.2 53.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
165 jondurbin airoboros-70b-3.3 53.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
166 Sakalti Saka-7.6B 53.8% base merges and moerges; float16 Official leaderboard
167 RDson WomboCombo-R1-Coder-14B-Preview 53.8% base merges and moerges; bfloat16 Official leaderboard
168 c4ai-command-r-08-2024 53.7% chat models (RLHF, DPO, IFT, ...); float16 Aug 2024 Official leaderboard
169 mistralai/Codestral-22B-v0.1 53.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
170 airev-ai Jais-Inception-70b 53.7% ; bfloat16 Official leaderboard
171 DeepSeek-R1-Distill-Llama-70B 53.6% fine-tuned on domain-specific datasets; bfloat16 Jan 2025 Official leaderboard
172 arcee-ai Arcee-Spark 53.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
173 Daemontatox Llama3.3-70B-CogniLink 53.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
174 huihui-ai Qwen2.5-7B-Instruct-abliterated 53.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
175 mssqpi ElmGemma2 53.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
176 Isaak-Carter Josiefied-Qwen2.5-7B-Instruct-abliterated-v2 53.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
177 Qwen2.5-14B 53.4% pretrained; bfloat16 Sep 2024 Official leaderboard
178 Cran-May T.E-8.1 53.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
179 dfurman Llama-3-70B-Orpo-v0.1 53.3% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
180 Aya Expanse 8B 53.3% chat models (RLHF, DPO, IFT, ...); float16 Oct 2024 Official leaderboard
181 migtissera Tess-3-Llama-3.1-70B 53.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
182 Qwen2.5-Math-72B-Instruct 53.1% fine-tuned on domain-specific datasets; bfloat16 Sep 2024 Official leaderboard
183 paloalma ECE-TW3-JRGL-V2 53.1% base merges and moerges; bfloat16 Official leaderboard
184 dnhkng RYS-Llama3.1-Large 53.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
185 vicgalle CarbonBeagle-11B 53% base merges and moerges; float16 Official leaderboard
186 inceptionai jais-family-13b-chat 52.9% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
187 Qwen/Qwen2-MoE-57B-A14B-Instruct 52.9% ; bfloat16 Official leaderboard
188 SicariusSicariiStuff Qwen2.5-14B_Uncencored 52.8% fine-tuned on domain-specific datasets; float16 Official leaderboard
189 SeaLLMs SeaLLM-7B-v2.5 52.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
190 meta-llama/Meta-Llama-3-70B 52.7% pretrained; bfloat16 Official leaderboard
191 nitky RoguePlanet-DeepSeek-R1-Qwen-32B 52.6% base merges and moerges; bfloat16 Official leaderboard
192 meta-llama/Meta-Llama-3.1-70B 52.6% pretrained; bfloat16 Official leaderboard
193 anthracite-org magnum-v3-9b-chatml 52.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
194 Llama 3.1-70B 52.5% pretrained; bfloat16 Jul 2024 Official leaderboard
195 abacusai Smaug-Llama-3-70B-Instruct-32K 52.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
196 FreedomIntelligence AceGPT-v2-70B-Chat 52.4% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
197 informatiker Qwen2-7B-Instruct-abliterated 52.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
198 CohereForAI/aya-23-35B 52.3% fine-tuned on domain-specific datasets; float16 Official leaderboard
199 TeetouchQQ model_mergev2 52.3% base merges and moerges; bfloat16 Official leaderboard
200 abacusai Smaug-Llama-3-70B-Instruct 52.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
201 natong19 Qwen2-7B-Instruct-abliterated 52.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
202 xiangxinai Xiangxin-2XL-Chat-1048k-Chinese-Llama3-70B 52% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
203 OpenBuddy openbuddy-llama3-70b-v21.2-32k 52% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
204 SicariusSicariiStuff Qwen2.5-14B_Uncensored_Instruct 51.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
205 failspy Smaug-Llama-3-70B-Instruct-abliterated-v3 51.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
206 vicgalle ConfigurableBeagle-11B 51.7% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
207 Qwen2-7B-Instruct 51.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Jun 2024 Official leaderboard
208 DeepMount00 Llama-3.1-8b-Ita 51.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
209 arcee-ai Llama-3.1-SuperNova-Lite 51.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
210 nyunai nyun-c2-llama3-56B 51.6% fine-tuned on domain-specific datasets; float16 Official leaderboard
211 princeton-nlp Llama-3-Base-8B-SFT-SimPO 51.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
212 paloalma ECE-TW3-JRGL-V1 51.4% base merges and moerges; float16 Official leaderboard
213 VAGOsolutions Llama-3.1-SauerkrautLM-8b-Instruct 51.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
214 Qwen/Qwen2.5-Math-72B 51.4% pretrained; bfloat16 Official leaderboard
215 vicgalle CarbonBeagle-11B-truthy 51.3% fine-tuned on domain-specific datasets; float16 Official leaderboard
216 NaniDAO deepseek-r1-qwen-2.5-32B-ablated 51.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
217 utter-project EuroLLM-9B-Instruct 51.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
218 elmrc ElmJuhaina70B 51.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
219 core42 jais-30b-chat-v1 51.1% ; ? Official leaderboard
220 nbeerbower DeepSeek-R1-Qwen-lorablated-32B 51% base merges and moerges; bfloat16 Official leaderboard
221 princeton-nlp Llama-3-Base-8B-SFT-IPO 50.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
222 FreedomIntelligence AceGPT-v2-32B 50.8% ; float16 Official leaderboard
223 nicoboss DeepSeek-R1-Distill-Qwen-32B-Uncensored 50.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
224 shadowml Mixolar-4x7b 50.8% base merges and moerges; float16 Official leaderboard
225 asas-ai AceGPT-v2-32B 50.8% fine-tuned on domain-specific datasets; float16 Official leaderboard
226 axolotl-ai-co romulus-mistral-nemo-12b-simpo 50.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
227 Qwen1.5-110B 50.6% pretrained; bfloat16 Apr 2024 Official leaderboard
228 invalid-coder Sakura-SOLAR-Instruct-CarbonVillain-en-10.7B-v2-slerp 50.5% base merges and moerges; float16 Official leaderboard
229 jeonsworld CarbonVillain-en-10.7B-v4 50.4% base merges and moerges; float16 Official leaderboard
230 kekmodel StopCarbon-10.7B-v5 50.4% base merges and moerges; float16 Official leaderboard
231 Gemma 2 9B 50.4% ; ? Jun 2024 Official leaderboard
232 nbeerbower mistral-nemo-wissenschaft-12B 50.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
233 FreedomIntelligence AceGPT-v1.5-13B-Chat 50.3% ; ? Official leaderboard
234 microsoft/Phi-3.5-mini-instruct 50.2% ; bfloat16 Official leaderboard
235 FreedomIntelligence AceGPT-v2-8B-Chat 50.2% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
236 Sakalti Saka-3.8B 50.1% base merges and moerges; float16 Official leaderboard
237 openchat openchat-3.5-0106-gemma 50% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
238 prithivMLmods Qwen2.5-14B-DeepSeek-R1-1M 50% base merges and moerges; bfloat16 Official leaderboard
239 upstage SOLAR-10.7B-Instruct-v1.0 50% fine-tuned on domain-specific datasets; float16 Official leaderboard
240 vicgalle ConfigurableSOLAR-10.7B 50% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
241 migtissera Tess-v2.5-Gemma-2-27B-alpha 49.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
242 RDson CoderO1-DeepSeekR1-Coder-32B-Preview 49.9% base merges and moerges; bfloat16 Official leaderboard
243 ai21labs Jamba-v0.1 49.9% pretrained; bfloat16 Official leaderboard
244 anthracite-org magnum-v2.5-12b-kto 49.9% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
245 Sakalti Saka-24B 49.9% base merges and moerges; float16 Official leaderboard
246 NLPark Shi-Cis-Kestrel-uncensored 49.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
247 vicgalle Configurable-Hermes-2-Pro-Llama-3-8B 49.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
248 llm-mle ara_mle_pretrained_8b 49.8% ; bfloat16 Official leaderboard
249 llm-mle arb_mle_pretrained_8b 49.8% pretrained; bfloat16 Official leaderboard
250 arcee-ai Llama-Spark 49.7% ; ? Official leaderboard
251 princeton-nlp Llama-3-Base-8B-SFT-RDPO 49.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
252 MarinaraSpaghetti NemoReRemix-12B 49.7% base merges and moerges; bfloat16 Official leaderboard
253 Qwen/Qwen1.5-110B-Chat 49.7% pretrained; bfloat16 Official leaderboard
254 AALF gemma-2-27b-it-SimPO-37K-100steps 49.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
255 princeton-nlp Llama-3-Base-8B-SFT-DPO 49.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
256 Magpie-Align Llama-3-8B-Magpie-Align-v0.3 49.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
257 grimjim Llama-3-Instruct-8B-SPPO-Iter3-SimPO-merge 49.3% base merges and moerges; bfloat16 Official leaderboard
258 MaziyarPanahi calme-2.3-qwen2-7b 49.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
259 AALF gemma-2-27b-it-SimPO-37K 49.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
260 vilm Quyen-Pro-Max-v0.1 49.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
261 Dampfinchen Llama-3.1-8B-Ultra-Instruct 49.1% base merges and moerges; bfloat16 Official leaderboard
262 openchat openchat-3.6-8b-20240522 49.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
263 Daemontatox MawaredT1 49.1% ; bfloat16 Official leaderboard
264 core42 jais-13b-chat 49% ; ? Official leaderboard
265 saltlux luxia-21.4b-alignment-v1.2 49% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
266 RLHFlow LLaMA3-iterative-DPO-final 49% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
267 turboderp Cat-Llama-3-70B-instruct 49% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
268 inceptionai jais-adapted-70b 49% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
269 Eurdem Defne-llama3.1-8B 49% base merges and moerges; bfloat16 Official leaderboard
270 NousResearch Nous-Hermes-2-Mixtral-8x7B-DPO 48.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
271 cognitivecomputations WestLake-7B-v2-laser 48.9% ; float16 Official leaderboard
272 natong19 Mistral-Nemo-Instruct-2407-abliterated 48.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
273 elmrc juhaina 48.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
274 elmrc Juhaina28B 48.8% ; ? Official leaderboard
275 lightblue suzume-llama-3-8B-multilingual-orpo-borda-top75 48.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
276 huihui-ai DeepSeek-R1-Distill-Qwen-32B-abliterated 48.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
277 VAGOsolutions SauerkrautLM-Phi-3-medium 48.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
278 VAGOsolutions SauerkrautLM-Qwen-32b 48.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
279 princeton-nlp Llama-3-Instruct-8B-SimPO-v0.2 48.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
280 Qwen2.5-7B 48.6% pretrained; bfloat16 Sep 2024 Official leaderboard
281 lemon-mint gemma-ko-7b-instruct-v0.62 48.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
282 princeton-nlp Llama-3-Base-8B-SFT-KTO 48.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
283 Danielbrdz Barcenas-14b-Phi-3-medium-ORPO 48.5% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
284 elmrc Juhaina29B 48.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
285 Mihaiii Llama-3-pruned-45B-Drobeta-Turnu-Severin 48.5% ; bfloat16 Official leaderboard
286 princeton-nlp Llama-3-8B-ProLong-64k-Instruct 48.5% ; ? Official leaderboard
287 v000000 L3.1-Niitorm-8B-DPO-t0.0001 48.5% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
288 MaziyarPanahi Llama-3-8B-Instruct-v0.10 48.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
289 princeton-nlp Llama-3-8B-ProLong-512k-Instruct 48.3% ; ? Official leaderboard
290 TIGER-Lab MAmmoTH2-8x7B-Plus 48.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
291 MaziyarPanahi calme-2.5-qwen2-7b 48.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
292 Qwen1.5-72B 48.3% pretrained; bfloat16 Feb 2024 Official leaderboard
293 jondurbin bagel-8b-v1.0 48.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
294 GritLM GritLM-8x7B-KTO 48.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
295 inceptionai jais-family-6p7b-chat 48.2% ; ? Official leaderboard
296 CohereForAI/aya-23-8B 48.1% fine-tuned on domain-specific datasets; float16 Official leaderboard
297 mlabonne NeuralDaredevil-8B-abliterated 48% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
298 MaziyarPanahi Calme-4x7B-MoE-v0.1 48% base merges and moerges; bfloat16 Official leaderboard
299 meta-llama/Meta-Llama-3.1-8B-Instruct 48% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
300 DeepSeek-R1-Distill-Qwen-14B 48% fine-tuned on domain-specific datasets; bfloat16 Jan 2025 Official leaderboard
301 OpenBuddy openbuddy-deepseek-67b-v18.1-4k 48% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
302 HeshamHaroon Testv1 48% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
303 MaziyarPanahi Calme-4x7B-MoE-v0.2 47.9% base merges and moerges; bfloat16 Official leaderboard
304 zhengr MixTAO-7Bx2-MoE-v8.1 47.9% base merges and moerges; bfloat16 Official leaderboard
305 ibivibiv multimaster-7b-v6 47.9% ; ? Official leaderboard
306 inceptionai jais-family-2p7b-chat 47.9% ; ? Official leaderboard
307 Kukedlc NeuralLLaMa-3-8b-DT-v0.1 47.8% base merges and moerges; float16 Official leaderboard
308 princeton-nlp Llama-3-Instruct-8B-RDPO 47.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
309 OpenBuddy openbuddy-qwen2.5llamaify-7b-v23.1-200k 47.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
310 airev-ai Jais-Inception-7b-V0.1 47.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
311 jsfs11 MixtureofMerges-MoE-4x7b-v5 47.8% base merges and moerges; bfloat16 Official leaderboard
312 kevin009 llamaRAGdrama 47.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
313 Kukedlc NeuralSynthesis-7B-v0.1 47.8% base merges and moerges; bfloat16 Official leaderboard
314 vicgalle Configurable-Hermes-3-Llama-3.1-8B 47.7% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
315 AIDC-AI Marco-LLM-AR 47.7% pretrained; bfloat16 Official leaderboard
316 princeton-nlp Llama-3-Instruct-8B-SimPO 47.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
317 Kukedlc NeuralSynthesis-7B-v0.3 47.7% base merges and moerges; bfloat16 Official leaderboard
318 saltlux luxia-21.4b-alignment-v1.0 47.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
319 MaziyarPanahi Llama-3-8B-Instruct-v0.9 47.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
320 nbeerbower gemma2-gutenberg-27B 47.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
321 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter3 47.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
322 airev-ai Jais-Inception-7b-V0.2 47.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
323 allknowingroger MultiverseEx26-7B-slerp 47.5% base merges and moerges; bfloat16 Official leaderboard
324 Qwen/Qwen1.5-32B-Chat 47.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
325 liminerity M7-7b 47.5% base merges and moerges; bfloat16 Official leaderboard
326 princeton-nlp Llama-3-Base-8B-SFT-SLiC-HF 47.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
327 grimjim Llama-3.1-8B-Instruct-abliterated_via_adapter 47.4% base merges and moerges; bfloat16 Official leaderboard
328 inceptionai jais-adapted-13b-chat 47.4% ; ? Official leaderboard
329 elmrc Juhaina-Instruct-v1 47.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
330 utter-project EuroLLM-9B 47.4% pretrained; bfloat16 Official leaderboard
331 jsfs11 MixtureofMerges-MoE-4x7b-v4 47.4% base merges and moerges; bfloat16 Official leaderboard
332 RubielLabarta LogoS-7Bx2-MoE-13B-v0.2 47.3% base merges and moerges; bfloat16 Official leaderboard
333 inceptionai jais-family-30b-16k-chat 47.3% ; ? Official leaderboard
334 ClusterlabAi GemmAr-7B-v1 47.3% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
335 princeton-nlp Llama-3-Base-8B-SFT-ORPO 47.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
336 NousResearch Hermes-3-Llama-3.1-8B 47.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
337 EpistemeAI Fireball-Alpaca-Llama-3.1-8B-Philos-DPO-200 47.3% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
338 meraGPT mera-mix-4x7B 47.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
339 mlabonne Daredevil-8B-abliterated 47.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
340 OpenBuddy openbuddy-qwen2.5llamaify-14b-v23.1-200k 47.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
341 sambanovasystems SambaLingo-Arabic-Base-70B 47.2% ; ? Official leaderboard
342 dzakwan dzakwan-MoE-4x7b-Beta 47.1% fine-tuned on domain-specific datasets; float16 Official leaderboard
343 cognitivecomputations laserxtral 47.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
344 Orenguteng Llama-3.1-8B-Lexi-Uncensored 47.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
345 TIGER-Lab MAmmoTH2-8B-Plus 47.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
346 Qwen1.5-14B-Chat 47.1% pretrained; bfloat16 Feb 2024 Official leaderboard
347 Kquant03 CognitiveFusion2-4x7B-BF16 47.1% base merges and moerges; bfloat16 Official leaderboard
348 MaziyarPanahi Topxtral-4x7B-v0.1 47.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
349 vicgalle Configurable-Llama-3.1-8B-Instruct 47.1% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
350 thariqkhalid ara_mle_sft_300_8b 47.1% ; bfloat16 Official leaderboard
351 princeton-nlp Llama-3-Instruct-8B-RDPO-v0.2 47.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
352 migtissera Tess-3-Mistral-Nemo-12B 47% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
353 Ali-C137 Ge1H10M-0000 47% pretrained; bfloat16 Official leaderboard
354 FreedomIntelligence AceGPT-13B-chat 47% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
355 princeton-nlp Llama-3-Instruct-8B-DPO-v0.2 47% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
356 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter2 47% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
357 Slim205 Barka-2b-it 47% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
358 EpistemeAI Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R2 47% fine-tuned on domain-specific datasets; float16 Official leaderboard
359 princeton-nlp Llama-3-Instruct-8B-DPO 46.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
360 bunnycore HyperLlama-3.1-8B 46.9% base merges and moerges; bfloat16 Official leaderboard
361 Kukedlc NeuralExperiment-7b-MagicCoder-v7.5 46.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
362 Orenguteng Llama-3.1-8B-Lexi-Uncensored-V2 46.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
363 EpistemeAI Fireball-Alpaca-Llama3.1.08-8B-Philos-C-R1-KTO-beta 46.9% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
364 MTSAIR multi_verse_model 46.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
365 Kukedlc NeuralSynthesis-7b-v0.4-slerp 46.9% base merges and moerges; bfloat16 Official leaderboard
366 lightblue suzume-llama-3-8B-multilingual-orpo-borda-full 46.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
367 NousResearch Nous-Hermes-2-SOLAR-10.7B 46.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
368 automerger YamshadowExperiment28-7B 46.8% base merges and moerges; bfloat16 Official leaderboard
369 princeton-nlp Llama-3-Instruct-8B-RRHF-v0.2 46.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
370 EpistemeAI2 Fireball-Alpaca-Llama-3.1-8B-Instruct-KTO-beta 46.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
371 NousResearch Hermes-2-Theta-Llama-3-8B 46.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
372 VAGOsolutions SauerkrautLM-Gemma-7b 46.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
373 princeton-nlp Llama-3-Instruct-8B-IPO 46.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
374 Kukedlc NeuralLLaMa-3-8b-ORPO-v0.3 46.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
375 EpistemeAI Fireball-Alpaca-Llama3.1.07-8B-Philos-Math-KTO-beta 46.7% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
376 VAGOsolutions Llama-3-SauerkrautLM-8b-Instruct 46.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
377 yunconglong Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B 46.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
378 allknowingroger Ph3della5-14B 46.7% base merges and moerges; float16 Official leaderboard
379 mlabonne Beyonder-4x7B-v3 46.7% base merges and moerges; float16 Official leaderboard
380 princeton-nlp Llama-3-Instruct-8B-SLiC-HF-v0.2 46.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
381 v000000 SwallowMaid-8B-L3-SPPO-abliterated 46.7% ; ? Official leaderboard
382 princeton-nlp Llama-3-Instruct-8B-IPO-v0.2 46.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
383 DeepMount00 Llama-3-8b-Ita 46.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
384 princeton-nlp Llama-3-Instruct-8B-ORPO-v0.2 46.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
385 jpacifico Chocolatine-14B-Instruct-DPO-v1.2 46.6% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
386 mlabonne ChimeraLlama-3-8B-v3 46.6% base merges and moerges; float16 Official leaderboard
387 Weyaxi Einstein-v7-Qwen2-7B 46.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
388 MaziyarPanahi calme-2.4-qwen2-7b 46.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
389 MaziyarPanahi calme-2.6-qwen2-7b 46.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
390 Qwen/Qwen2.5-Coder-7B-Instruct 46.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
391 OpenBuddy openbuddy-qwen1.5-32b-v21.1-32k 46.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
392 Qwen/Qwen2-MoE-57B-A14B 46.5% ; bfloat16 Official leaderboard
393 yunconglong MoE_13B_DPO 46.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
394 cognitivecomputations dolphin-2.9.2-qwen2-7b 46.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
395 MaziyarPanahi calme-2.2-qwen2-7b 46.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
396 sophosympatheia Midnight-Miqu-70B-v1.5 46.5% base merges and moerges; float16 Official leaderboard
397 MaziyarPanahi calme-2.1-qwen2-7b 46.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
398 inceptionai jais-adapted-7b-chat 46.5% ; ? Official leaderboard
399 Gemma 1.1 7B IT 46.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Apr 2024 Official leaderboard
400 elmrc JuhainaL2 46.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
401 uygarkurt llama-3-merged-linear 46.4% base merges and moerges; float16 Official leaderboard
402 shadowml BeagSake-7B 46.4% base merges and moerges; bfloat16 Official leaderboard
403 jpacifico Chocolatine-14B-Instruct-4k-DPO 46.4% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
404 mlabonne AlphaMonarch-7B 46.4% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
405 OpenBuddy openbuddy-qwen1.5-32b-v21.2-32k 46.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
406 Qwen/Qwen2.5-3B-Instruct 46.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
407 princeton-nlp Llama-3-Instruct-8B-KTO-v0.2 46.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
408 elmrc JuhainaL 46.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
409 MaziyarPanahi Llama-3-8B-Instruct-v0.8 46.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
410 phi-3-medium 14B 46.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Apr 2024 Official leaderboard
411 cstr llama3-8b-spaetzle-v37 46.2% base merges and moerges; bfloat16 Official leaderboard
412 BAAI Infinity-Instruct-3M-0625-Llama3-8B 46.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
413 mlabonne NeuralMonarch-7B 46.1% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
414 yunconglong DARE_TIES_13B 46.1% base merges and moerges; bfloat16 Official leaderboard
415 MaziyarPanahi calme-2.7-qwen2-7b 46.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
416 CultriX NeuralMona_MoE-4x7B 46.1% base merges and moerges; bfloat16 Official leaderboard
417 grimjim Llama-3-Instruct-8B-SimPO-SPPO-Iter3-merge 46.1% base merges and moerges; bfloat16 Official leaderboard
418 SeaLLMs SeaLLMs-v3-7B-Chat 46.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
419 UCLA-AGI Llama-3-Instruct-8B-SPPO-Iter1 46.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
420 Qwen2.5-Coder-14B-Instruct 46.1% fine-tuned on domain-specific datasets; bfloat16 Nov 2024 Official leaderboard
421 dnhkng RYS-Llama-3.1-8B-Instruct 46% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
422 Gemma 7B 46% pretrained; bfloat16 Feb 2024 Official leaderboard
423 inceptionai jais-family-30b-16k 46% ; ? Official leaderboard
424 Qwen2.5-Coder-14B 46% pretrained; bfloat16 Sep 2024 Official leaderboard
425 macadeliccc Samantha-Qwen-2-7B 45.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
426 freewheelin free-solar-evo-v0.13 45.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
427 leafspark Llama-3.1-8B-MultiReflection-Instruct 45.9% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
428 mlabonne ChimeraLlama-3-8B-v2 45.8% base merges and moerges; float16 Official leaderboard
429 Magpie-Align MagpieLM-8B-Chat-v0.1 45.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
430 maywell Qwen2-7B-Multilingual-RP 45.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
431 princeton-nlp Llama-3-Instruct-8B-KTO 45.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
432 inceptionai jais-family-30b-8k 45.8% ; ? Official leaderboard
433 princeton-nlp Llama-3-Instruct-8B-CPO-v0.2 45.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
434 BAAI Infinity-Instruct-7M-Gen-Llama3_1-8B 45.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
435 wenbopan Faro-Yi-9B-DPO 45.8% ; bfloat16 Official leaderboard
436 freewheelin free-solar-evo-v0.11 45.8% fine-tuned on domain-specific datasets; float16 Official leaderboard
437 Qwen1.5-32B 45.8% pretrained; bfloat16 Feb 2024 Official leaderboard
438 Phi-3-medium-128k-instruct 45.7% chat models (RLHF, DPO, IFT, ...); bfloat16 May 2024 Official leaderboard
439 lightblue suzume-llama-3-8B-multilingual-orpo-borda-top25 45.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
440 EpistemeAI Fireball-Meta-Llama-3.1-8B-Instruct-Agent-0.003-128K-code 45.7% fine-tuned on domain-specific datasets; float16 Official leaderboard
441 gaverfraxz Meta-Llama-3.1-8B-Instruct-HalfAbliterated-TIES 45.6% base merges and moerges; float16 Official leaderboard
442 Weyaxi Bagel-Hermes-34B-Slerp 45.6% base merges and moerges; bfloat16 Official leaderboard
443 Kukedlc NeuralLLaMa-3-8b-ORPO-v0.4 45.6% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
444 HeshamHaroon llama-3-instruct-slerp-arabic 45.6% ; bfloat16 Official leaderboard
445 freewheelin free-solar-evo-v0.1 45.5% fine-tuned on domain-specific datasets; float16 Official leaderboard
446 Ali-C137 Q2AW1M-0000 45.5% pretrained; bfloat16 Official leaderboard
447 Ali-C137 Q2AW1M-1000 45.5% pretrained; bfloat16 Official leaderboard
448 Ali-C137 Q2AW1M-1010 45.5% pretrained; bfloat16 Official leaderboard
449 jondurbin bagel-dpo-34b-v0.2 45.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
450 Ali-C137 PG7BB 45.5% pretrained; bfloat16 Official leaderboard
451 Ali-C137 Q2H10M-0000 45.5% pretrained; bfloat16 Official leaderboard
452 abacusai Smaug-Mixtral-v0.1 45.5% ; ? Official leaderboard
453 Danielbrdz Barcenas-Llama3-8b-ORPO 45.5% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
454 Ali-C137 Q2AW1M-0010 45.5% pretrained; bfloat16 Official leaderboard
455 Ali-C137 Q2AW1M-1100 45.5% pretrained; bfloat16 Official leaderboard
456 abacusai bigstral-12b-32k 45.4% base merges and moerges; float16 Official leaderboard
457 chujiezheng Mistral7B-PairRM-SPPO-ExPO 45.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
458 cloudyu TomGrc_FusionNet_34Bx2_MoE_v0.1_DPO_f16 45.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
459 mssqpi ALQLora2 45.4% ; ? Official leaderboard
460 Ali-C137 Q2AW1M-0100 45.4% pretrained; bfloat16 Official leaderboard
461 ibivibiv llama-3-nectar-dpo-8B 45.3% ; ? Official leaderboard
462 2A2I-R L500MT 45.3% pretrained; bfloat16 Official leaderboard
463 2A2I-R L100MT 45.3% pretrained; bfloat16 Official leaderboard
464 cloudyu Mixtral_34Bx2_MoE_60B 45.3% base merges and moerges; bfloat16 Official leaderboard
465 meta-llama/Meta-Llama-3-8B-Instruct 45.3% fine-tuned on domain-specific datasets; float16 Official leaderboard
466 Magpie-Align Llama-3-8B-Magpie-Align-SFT-v0.3 45.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
467 2A2I-R L200MT 45.3% pretrained; bfloat16 Official leaderboard
468 2A2I-R L1000MT 45.2% pretrained; bfloat16 Official leaderboard
469 Magpie-Align Llama-3.1-8B-Magpie-Align-SFT-v0.1 45.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
470 core42 jais-30b-v3 45.2% ; ? Official leaderboard
471 GritLM GritLM-8x7B 45.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
472 cloudyu Yi-34Bx2-MOE-200K 45.1% base merges and moerges; bfloat16 Official leaderboard
473 UCLA-AGI Mistral7B-PairRM-SPPO 45.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
474 UCLA-AGI Mistral7B-PairRM-SPPO-Iter2 45.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
475 Magpie-Align MagpieLM-8B-SFT-v0.1 45.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
476 princeton-nlp Llama-3-Base-8B-SFT-CPO 45% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
477 BAAI Infinity-Instruct-3M-0625-Qwen2-7B 45% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
478 abacusai Slerp-CM-mist-dpo 45% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
479 princeton-nlp Mistral-7B-Instruct-RRHF 44.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
480 failspy Phi-3-medium-4k-instruct-abliterated-v3 44.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
481 cloudyu Yi-34Bx2-MoE-60B 44.9% base merges and moerges; bfloat16 Official leaderboard
482 UCLA-AGI Mistral7B-PairRM-SPPO-Iter3 44.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
483 UCLA-AGI Mistral7B-PairRM-SPPO-Iter1 44.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
484 princeton-nlp Llama-3-Base-8B-SFT-RRHF 44.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
485 nvidia Llama3-ChatQA-1.5-8B 44.8% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
486 princeton-nlp Llama-3-Instruct-8B-RRHF 44.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
487 princeton-nlp Mistral-7B-Instruct-SimPO 44.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
488 princeton-nlp Llama-3-Instruct-8B-CPO 44.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
489 failspy Codestral-22B-v0.1-abliterated-v3 44.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
490 silma-ai SILMA-Kashif-2B-Instruct-v1.0 44.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
491 cloudyu Yi-34Bx2-MoE-60B-DPO 44.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
492 vicgalle Configurable-Llama-3-8B-v0.3 44.6% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
493 princeton-nlp Mistral-7B-Instruct-DPO 44.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
494 cognitivecomputations dolphin-2.9.3-llama-3-8b 44.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
495 cognitivecomputations dolphin-2.9.4-llama3.1-8b 44.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
496 cognitivecomputations Llama-3-8B-Instruct-abliterated-v2 44.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
497 cognitivecomputations dolphin-2.6-mixtral-8x7b 44.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
498 freewheelin free-llama3-dpo-v0.2 44.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
499 Ali-C137 Q2AW1M-1001 44.3% pretrained; bfloat16 Official leaderboard
500 princeton-nlp Llama-3-Instruct-8B-SLiC-HF 44.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
501 01-ai/Yi-Coder-9B-Chat 44.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
502 Ali-C137 Q2AW1M-0001 44.3% pretrained; bfloat16 Official leaderboard
503 Artples L-MChat-7b 44.3% base merges and moerges; bfloat16 Official leaderboard
504 cognitivecomputations dolphin-2.7-mixtral-8x7b 44.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
505 ConvexAI Luminex-34B-v0.2 44.2% ; bfloat16 Official leaderboard
506 abacusai Smaug-34B-v0.1 44.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
507 EpistemeAI2 Fireball-Llama-3.1-8B-Philos-Relection 44.1% fine-tuned on domain-specific datasets; float16 Official leaderboard
508 princeton-nlp Llama-3-Base-8B-SFT 44.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
509 mlabonne Meta-Llama-3.1-8B-Instruct-abliterated 44.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
510 OpenBuddy openbuddy-llama3.1-8b-v22.3-131k 44.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
511 princeton-nlp Llama-3-Instruct-8B-ORPO 44.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
512 fblgit UNA-SimpleSmaug-34b-v1beta 44% base merges and moerges; bfloat16 Official leaderboard
513 Omartificial-Intelligence-Space al-baka-llama3-8b-experimental 44% fine-tuned on domain-specific datasets; float16 Official leaderboard
514 lightblue suzume-llama-3-8B-multilingual-orpo-borda-half 44% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
515 princeton-nlp Mistral-7B-Base-SFT-KTO 44% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
516 vicgalle Configurable-Mistral-7B 44% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
517 MaziyarPanahi Mistral-7B-Instruct-v0.3 43.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
518 GritLM GritLM-7B-KTO 43.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
519 princeton-nlp Mistral-7B-Instruct-RDPO 43.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
520 Qwen2-7B 43.8% pretrained; bfloat16 Jun 2024 Official leaderboard
521 ConvexAI Luminex-34B-v0.1 43.8% ; bfloat16 Official leaderboard
522 mistralai/Mistral-7B-Instruct-v0.3 43.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
523 princeton-nlp Mistral-7B-Instruct-CPO 43.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
524 tiiuae/falcon-180B-chat 43.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
525 inceptionai jais-family-1p3b-chat 43.8% ; ? Official leaderboard
526 princeton-nlp Llama-3-8B-ProLong-64k-Base 43.7% ; ? Official leaderboard
527 princeton-nlp Mistral-7B-Instruct-KTO 43.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
528 cognitivecomputations openchat-3.5-0106-laser 43.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
529 openchat openchat-3.5-0106 43.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
530 Ali-C137 Q2AW1M-0111 43.7% pretrained; bfloat16 Official leaderboard
531 arcee-ai Virtuoso-Lite 43.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
532 rombodawg rombos_Replete-Coder-Instruct-8b-Merged 43.6% base merges and moerges; bfloat16 Official leaderboard
533 Ali-C137 Q2AW1M-1111 43.6% pretrained; bfloat16 Official leaderboard
534 Qwen2.5-3B 43.6% pretrained; bfloat16 Sep 2024 Official leaderboard
535 cognitivecomputations dolphin-2.9.1-llama-3-8b 43.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
536 wenbopan Faro-Yi-34B-DPO 43.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
537 Mistral-7B-Instruct-v0.2 43.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Dec 2023 Official leaderboard
538 dfurman Llama-3-8B-Orpo-v0.1 43.5% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
539 failspy Meta-Llama-3-8B-Instruct-abliterated-v3 43.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
540 ibivibiv llama3-8b-instruct-summary-combined 43.4% ; ? Official leaderboard
541 Qwen2.5-Coder (7B) 43.3% pretrained; bfloat16 Sep 2024 Official leaderboard
542 lightblue suzume-llama-3-8B-multilingual 43.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
543 jondurbin bagel-dpo-34b-v0.5 43.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
544 01-ai/Yi-1.5-34B-Chat-16K 43.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
545 Qwen1.5-7B-Chat 43.1% pretrained; bfloat16 Feb 2024 Official leaderboard
546 INSAIT-Institute BgGPT-7B-Instruct-v0.2 43.1% pretrained; bfloat16 Official leaderboard
547 Ali-C137 L3H10M-0000 43% pretrained; bfloat16 Official leaderboard
548 ValiantLabs Llama3.1-8B-ShiningValiant2 43% ; ? Official leaderboard
549 BAAI Infinity-Instruct-3M-0625-Yi-1.5-9B 43% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
550 Llama 3.1-8B 43% pretrained; bfloat16 Jul 2024 Official leaderboard
551 princeton-nlp Mistral-7B-Instruct-IPO 43% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
552 meta-llama/Meta-Llama-3.1-8B 42.9% pretrained; bfloat16 Official leaderboard
553 vicgalle ConfigurableHermes-7B 42.9% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
554 abacusai Llama-3-Smaug-8B 42.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
555 tokyotech-llm Llama-3-Swallow-8B-Instruct-v0.1 42.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
556 anthracite-org magnum-v3-34b 42.9% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
557 migtissera Tess-v2.5-Phi-3-medium-128k-14B 42.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
558 meta-llama/Llama-3.2-3B-Instruct 42.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
559 beowolx CodeNinja-1.0-OpenChat-7B 42.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
560 CombinHorizon YiSM-blossom5.1-34B-SLERP 42.6% base merges and moerges; bfloat16 Official leaderboard
561 princeton-nlp Mistral-7B-Instruct-ORPO 42.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
562 wenbopan Faro-Yi-9B 42.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
563 cognitivecomputations dolphin-2.9.1-yi-1.5-34b 42.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
564 SeaLLMs SeaLLM-7B-v2 42.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
565 BAAI Infinity-Instruct-3M-0613-Mistral-7B 42.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
566 altomek YiSM-34B-0rn 42.4% base merges and moerges; float16 Official leaderboard
567 meta-llama/Meta-Llama-3-8B 42.4% pretrained; bfloat16 Official leaderboard
568 vicgalle Configurable-Janus-7B 42.3% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
569 terrycraddock Reflection-Llama-3.1-8B 42.3% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
570 FreedomIntelligence AceGPT-7B-chat 42.2% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
571 01-ai/Yi-Coder-9B 42.2% pretrained; bfloat16 Official leaderboard
572 vicgalle Configurable-Yi-1.5-9B-Chat 42.2% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
573 abacusai Liberated-Qwen1.5-14B 42.2% fine-tuned on domain-specific datasets; float16 Official leaderboard
574 upstage SOLAR-10.7B-v1.0 42.2% pretrained; float16 Official leaderboard
575 google/gemma-7b-it 42.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
576 Falcon3-10B-Instruct 42.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Dec 2024 Official leaderboard
577 princeton-nlp Mistral-7B-Instruct-SLiC-HF 42.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
578 01-ai/Yi-1.5-34B-Chat 42.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
579 01-ai/Yi-1.5-9B-Chat-16K 42% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
580 cognitivecomputations dolphin-2.9.3-mistral-7B-32k 42% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
581 Qwen1.5-14B 41.9% pretrained; bfloat16 Feb 2024 Official leaderboard
582 Yi-1.5-9B-Chat 41.9% chat models (RLHF, DPO, IFT, ...); bfloat16 May 2024 Official leaderboard
583 oobabooga CodeBooga-34B-v0.1 41.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
584 brucethemoose Yi-34B-200K-DARE-merge-v7 41.8% base merges and moerges; bfloat16 Official leaderboard
585 unsloth llama-3-8b-bnb-4bit 41.7% ; GPTQ Official leaderboard
586 oddadmix Arabic-Qwen2.5-3B 41.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
587 wenbopan Faro-Yi-34B 41.7% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
588 chujiezheng Smaug-34B-v0.1-ExPO 41.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
589 meta-llama/Llama-2-70b-hf 41.6% pretrained; float16 Official leaderboard
590 Weyaxi Nous-Hermes-2-SUS-Chat-34B-Slerp 41.5% base merges and moerges; bfloat16 Official leaderboard
591 royallab ZephRP-m7b 41.5% base merges and moerges; bfloat16 Official leaderboard
592 byroneverson Yi-1.5-9B-Chat-16K-abliterated 41.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
593 byroneverson Yi-1.5-9B-Chat-abliterated 41.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
594 Ali-C137 Q15H10M-0000 41.4% pretrained; bfloat16 Official leaderboard
595 cognitivecomputations dolphin-2.9.3-mistral-nemo-12b 41.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
596 sambanovasystems SambaLingo-Arabic-Chat 41.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
597 Falcon3-7B-Instruct 41.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Dec 2024 Official leaderboard
598 cognitivecomputations dolphin-2.9.1-yi-1.5-9b 41.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
599 Omartificial-Intelligence-Space Arabic-llama3.1-16bit-FT 41.1% ; bfloat16 Official leaderboard
600 speakleash Bielik-11B-v2.3-Instruct 41.1% base merges and moerges; float16 Official leaderboard
601 rombodawg Everyone-Coder-4x7b-Base 41.1% ; bfloat16 Official leaderboard
602 OpenBuddy openbuddy-zero-56b-v21.2-32k 41% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
603 Azure99 blossom-v5.1-34b 41% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
604 Falcon-180B 41% pretrained; bfloat16 Sep 2023 Official leaderboard
605 FreedomIntelligence AceGPT-13B 41% pretrained; float16 Official leaderboard
606 qnguyen3 Master-Yi-9B 41% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
607 FreedomIntelligence AceGPT-v1.5-13B 41% ; ? Official leaderboard
608 oddadmix Arabic-Qwen2.5-3B-v2.1 41% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
609 uukuguy speechless-code-mistral-7b-v1.0 40.9% fine-tuned on domain-specific datasets; float16 Official leaderboard
610 ghost-x ghost-8b-beta-1608 40.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
611 inceptionai jais-adapted-13b 40.8% ; ? Official leaderboard
612 618AI dictalm2-it-qa-fine-tune 40.8% ; bfloat16 Official leaderboard
613 deepseek-ai/DeepSeek-R1-Distill-Llama-8B 40.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
614 Qwen1.5-7B 40.7% pretrained; bfloat16 Feb 2024 Official leaderboard
615 Qwen/Qwen2.5-Math-7B 40.6% pretrained; bfloat16 Official leaderboard
616 rombodawg rombos_Replete-Coder-Llama3-8B 40.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
617 uukuguy speechless-zephyr-code-functionary-7b 40.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
618 OpenBuddy openbuddy-yi1.5-34b-v21.3-32k 40.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
619 princeton-nlp Mistral-7B-Base-SFT-SimPO 40.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
620 aws-prototyping MegaBeam-Mistral-7B-512k 40.4% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
621 cognitivecomputations dolphin-2.9.2-Phi-3-Medium 40.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
622 SakanaAI DiscoPOP-zephyr-7b-gemma 40.4% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
623 GritLM GritLM-7B 40.3% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
624 Qwen/Qwen2.5-1.5B-Instruct 40.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
625 Qwen2-1.5B-Instruct 40.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Jun 2024 Official leaderboard
626 MohamedRashad Arabic-Orpo-Llama-3-8B-Instruct 40.3% fine-tuned on domain-specific datasets; float16 Official leaderboard
627 speakleash Bielik-11B-v2 40.3% pretrained; bfloat16 Official leaderboard
628 kaist-ai janus-dpo-7b 40.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
629 princeton-nlp Mistral-7B-Base-SFT-RDPO 40.2% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
630 cognitivecomputations dolphin-2.9.2-Phi-3-Medium-abliterated 40.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
631 TIGER-Lab MAmmoTH2-7B-Plus 40.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
632 kaist-ai janus-7b 40.1% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
633 cognitivecomputations dolphin-2.9.3-Yi-1.5-34B-32k 40.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
634 kaist-ai mistral-orpo-capybara-7k 40% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
635 princeton-nlp Mistral-7B-Base-SFT-SLiC-HF 40% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
636 uukuguy zephyr-7b-alpha-dare-0.85 40% fine-tuned on domain-specific datasets; float16 Official leaderboard
637 cognitivecomputations dolphin-2.9.1-mixtral-1x22b 39.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
638 oddadmix Arabic-Qwen2.5-3B-v2 39.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
639 CausalLM 34b-beta 39.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
640 google/recurrentgemma-2b-it 39.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
641 amazon MegaBeam-Mistral-7B-300k 39.7% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
642 core42 jais-30b-v1 39.6% ; ? Official leaderboard
643 Qwen2.5-1.5B 39.6% pretrained; bfloat16 Sep 2024 Official leaderboard
644 mlabonne FineLlama-3.1-8B 39.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
645 01-ai/Yi-1.5-9B 39.5% pretrained; bfloat16 Official leaderboard
646 NLPark AnFeng_v3.1-Avocet 39.5% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
647 01-ai/Yi-1.5-34B-32K 39.5% pretrained; bfloat16 Official leaderboard
648 01-ai/Yi-1.5-9B-32K 39.5% pretrained; bfloat16 Official leaderboard
649 01-ai/Yi-9B-200K 39.5% pretrained; bfloat16 Official leaderboard
650 tiiuae/Falcon3-10B-Base 39.4% pretrained; bfloat16 Official leaderboard
651 maywell Synatra-7B-v0.3-RP 39.4% fine-tuned on domain-specific datasets; float16 Official leaderboard
652 Qwen/Qwen1.5-MoE-A2.7B 39.4% pretrained; bfloat16 Official leaderboard
653 Qwen/Qwen1.5-MoE-A2.7B-Chat 39.4% pretrained; bfloat16 Official leaderboard
654 princeton-nlp Mistral-7B-Base-SFT-RRHF 39.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
655 Ali-C137 M03H10M-0000 39.2% pretrained; bfloat16 Official leaderboard
656 Sakalti Saka-1.5B 39.1% base merges and moerges; float16 Official leaderboard
657 OpenBuddy openbuddy-yi1.5-34b-v21.6-32k-fp16 39.1% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
658 princeton-nlp Mistral-7B-Base-SFT-CPO 39% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
659 dicta-il dictalm2.0 39% pretrained; bfloat16 Official leaderboard
660 princeton-nlp Mistral-7B-Base-SFT-IPO 39% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
661 Yi-1.5-34B 38.9% pretrained; bfloat16 May 2024 Official leaderboard
662 NotAiLOL Yi-1.5-dolphin-9B 38.9% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
663 DeepSeek-R1-Distill-Qwen-7B 38.8% fine-tuned on domain-specific datasets; bfloat16 Jan 2025 Official leaderboard
664 Syed-Hasan-8503 Phi-3-mini-4K-instruct-cpo-simpo 38.8% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
665 Mistral 7B v0.3 38.7% pretrained; bfloat16 Sep 2023 Official leaderboard
666 MaziyarPanahi Mistral-7B-v0.3 38.7% pretrained; bfloat16 Official leaderboard
667 inceptionai jais-adapted-7b 38.6% ; ? Official leaderboard
668 dicta-il dictalm2.0-instruct 38.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
669 princeton-nlp Mistral-7B-Base-SFT-DPO 38.6% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
670 sambanovasystems SambaLingo-Arabic-Base 38.5% ; ? Official leaderboard
671 SenseLLM ReflectionCoder-CL-34B 38.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
672 inceptionai jais-family-13b 38.3% ; ? Official leaderboard
673 kaist-ai janus-orpo-7b 38.3% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
674 BAAI OPI-Llama-3.1-8B-Instruct 38.2% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
675 migtissera Tess-M-v1.3 38.2% fine-tuned on domain-specific datasets; float16 Official leaderboard
676 Gemma 1.1 2B IT 38.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Apr 2024 Official leaderboard
677 sajaw llama-2-7b-RandomGPT-5K-ar 38% ; bfloat16 Official leaderboard
678 FreedomIntelligence AceGPT-7B 38% pretrained; float16 Official leaderboard
679 abacusai bigyi-15b 38% base merges and moerges; float16 Official leaderboard
680 tiiuae/Falcon3-7B-Base 38% pretrained; bfloat16 Official leaderboard
681 core42 jais-13b 37.9% ; ? Official leaderboard
682 haoranxu ALMA-13B-Pretrain 37.9% ; ? Official leaderboard
683 Qwen2-1.5B 37.9% pretrained; bfloat16 Jun 2024 Official leaderboard
684 google/gemma-2b-it 37.8% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
685 Qwen/Qwen1.5-4B 37.7% pretrained; bfloat16 Official leaderboard
686 meta-llama/Llama-2-13b-hf 37.6% pretrained; float16 Official leaderboard
687 gpt2 37.6% ; GPTQ Official leaderboard
688 inceptionai jais-family-6p7b 37.6% ; ? Official leaderboard
689 Qwen/Qwen1.5-1.8B-Chat 37.5% pretrained; bfloat16 Official leaderboard
690 Qwen/Qwen2.5-Math-7B-Instruct 37.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
691 google/recurrentgemma-2b 37.4% ; ? Official leaderboard
692 Qwen1.5-4B-Chat 37.3% pretrained; bfloat16 Feb 2024 Official leaderboard
693 Ali-C137 L2H10M-0000 37.3% pretrained; bfloat16 Official leaderboard
694 meta-llama/Llama-2-7b-hf 37.2% pretrained; float16 Official leaderboard
695 tiiuae/Falcon3-1B-Base 37% pretrained; bfloat16 Official leaderboard
696 SenseLLM ReflectionCoder-DS-33B 37% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
697 inceptionai jais-family-590m-chat 36.8% ; ? Official leaderboard
698 Qwen/Qwen1.5-1.8B 36.7% pretrained; bfloat16 Official leaderboard
699 haoranxu ALMA-13B-R 36.6% chat models (RLHF, DPO, IFT, ...); float16 Official leaderboard
700 Qwen2-0.5B-Instruct 36.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Jun 2024 Official leaderboard
701 byroneverson Yi-1.5-34B-Chat-abliterated 36.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
702 NTQAI Nxcode-CQ-7B-orpo 36.5% chat models (RLHF, DPO, IFT, ...); bfloat16 Official leaderboard
703 Qwen2.5-0.5B 36.4% pretrained; bfloat16 Sep 2024 Official leaderboard
704 inceptionai jais-family-2p7b 36.3% ; ? Official leaderboard
705 Qwen/Qwen1.5-0.5B-Chat 36.2% pretrained; bfloat16 Official leaderboard
706 Llama 3.2 1B 36.2% pretrained; bfloat16 Sep 2024 Official leaderboard
707 haoranxu ALMA-7B-R 36.1% ; float16 Official leaderboard
708 Falcon3-3B-Instruct 36.1% chat models (RLHF, DPO, IFT, ...); bfloat16 Dec 2024 Official leaderboard
709 Falcon3-1B-Instruct 36% chat models (RLHF, DPO, IFT, ...); bfloat16 Dec 2024 Official leaderboard
710 inceptionai jais-family-1p3b 36% ; ? Official leaderboard
711 Gemma 2B 36% pretrained; bfloat16 Feb 2024 Official leaderboard
712 riotu-lab ArabianGPT-0.3B-QA 36% ; ? Official leaderboard
713 inceptionai jais-family-590m 35.9% ; ? Official leaderboard
714 bigscience bloom-7b1 35.9% pretrained; float16 Official leaderboard
715 tiiuae/falcon-11B 35.7% pretrained; bfloat16 Official leaderboard
716 Qwen/Qwen1.5-0.5B 35.7% pretrained; bfloat16 Official leaderboard
717 riotu-lab ArabianGPT-01B 35.7% ; ? Official leaderboard
718 riotu-lab ArabianGPT-03B 35.5% ; ? Official leaderboard
719 tiiuae/falcon-7b-instruct 35.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
720 Falcon-40B 35.5% pretrained; bfloat16 Mar 2023 Official leaderboard
721 riotu-lab ArabianGPT-08B-V2 35.4% ; ? Official leaderboard
722 kevinpro Hydra-LLaMA3-8B-v0.1 35.4% ; ? Official leaderboard
723 Qwen2-0.5B 35.3% pretrained; bfloat16 Jun 2024 Official leaderboard
724 Artples L-MChat-Small 35.2% base merges and moerges; bfloat16 Official leaderboard
725 tiiuae/Falcon3-3B-Base 35.2% pretrained; bfloat16 Official leaderboard
726 kevinpro Hydra-LLaMA3-8B-v0.2 35.1% ; ? Official leaderboard
727 Falcon-7B 35% pretrained; bfloat16 Apr 2023 Official leaderboard
728 Ali-C137 F1H10M-0000 35% pretrained; bfloat16 Official leaderboard
729 paloalma ECE-TW3-JRGL-V3 35% base merges and moerges; bfloat16 Official leaderboard
730 paloalma ECE-TW3-JRGL-V4 34.8% base merges and moerges; bfloat16 Official leaderboard
731 riotu-lab ArabianGPT-1.5B 34.7% ; ? Official leaderboard
732 tiiuae/falcon-40b-instruct 34.6% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
733 kaist-ai janus-rm-7b 34.5% fine-tuned on domain-specific datasets; bfloat16 Official leaderboard
734 MBZUAI MobiLlama-1B 34.4% ; ? Official leaderboard

3 models have held number one on the official leaderboard since its release in May 2024

Lab claims count from the day the card came out, independent runs from the model’s release.

Cohere Qwen
Own leaderboard Replaced by 2, Feb 2025
Qwen2.5-72B 69.1%

Source: Official Open Arabic LLM Leaderboard v1 leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2023 2024 2025 Near the ceiling (85%) Qwen2.5-72B Qwen2-72B Command R (03-2024) Qwen1.5-72B Mistral-7B-Instruct-v0.2 Falcon-40B 0% 20% 40% 60% 80% 100% 2023 2024 2025 Near the ceiling (85%) Qwen2.5-72B Command R (03-2024)

Source: Official Open Arabic LLM Leaderboard v1 leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

Versions

We switch the top row to each new version when it comes out. Open it for one row per version. An old version stays in play while people still test it, so whoever first takes it past 85% gets credit for closing it out.

Cohere Qwen
Open Arabic LLM Leaderboard v2 Not tested since Aug 2025
Qwen2.5-72B 72.4%
Open Arabic LLM Leaderboard v1 Replaced by 2, Feb 2025
Qwen2.5-72B 69.1%
Open Arabic LLM Leaderboard v2 Not tested since Aug 2025
Qwen2.5-72B 72.4%

Source: Official leaderboards

Replaced by Open Arabic LLM Leaderboard v2

More in Multilingual

All 120 in Multilingual
  • MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
  • WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET. K-EXAONE 90.5%
  • MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
  • CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories. Solar Open 2 90.7%
  • Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Nemotron 3 Ultra 550B-A55B Base 90.1%
  • MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%