Benchmark Almanac

nocot-bench (NCRI)

GPT-6 Astra 173.9 leads since September 2026
GPT-6 Astra 173.9 leads since September 2026
Only in system cards

The official leaderboard hasn’t scored a model released after Tencent Hy4 (November 2026). Labs report their own numbers for one newer model: GPT-6.2 Astra at 171.0.

Reasoning problems answered without chain of thought (e.g., Brew multi-hop tasks). Measures forward-pass reasoning relevant to monitorability.

Released
September 2026
Kind
Deception and alignment
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Another benchmark
Run by
Anyone
In views
Loss of control
Strongest model tested
GPT-6.2 Astra #1 on the capability index
Last in a system card
November 2026
Last model added
November 2026
Scores refreshed
November 20th, 2026
287
Entries on its own leaderboard, 201 models we track
2
System cards report it, from 2 labs

GPT-6 Astra leads with 173.9 GPT-6 Astra leads with 173.9

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

GPT-6 Astra leads with 173.9, claims included

Best score per model from every source, including labs’ own numbers and claims Epoch AI collected from reports and papers. Labs pick their own settings, so a claim can beat an independent run of the same model. Striped bars are claims, and dotted bars are reruns by Epoch AI or another board.

287 entries. Scroll sideways for the rest.

  1. 173.9 GPT-6 Astra
  2. 156.6 GPT-6.1 Sol
  3. 154.1 Gemini 4 Argon
  4. 149.6 GPT-6.1 Terra
  5. 129.7 Gemini 3.8 Flash
  6. 129.6 Claude Fable 5.1
  7. 127.6 Gemini 3.7 Flash
  8. 123.9 Gemini 3.1 Pro
  9. 122.2 GPT-6 Sol
  10. 121.5 Claude Fable 5
  11. 119.5 GPT-5.6 Sol
  12. 118.9 Claude Opus 5.5
  13. 115.8 Claude Opus 5
  14. 111.0 GPT-5.5
  15. 110.9 Gemini 3.6 Flash
  16. 110.8 Claude Opus 4.6
  17. 110.5 Gemini 3 Flash
  18. 110.5 Claude Opus 4.5
  19. 109.0 Claude Opus 4
  20. 108.8 Muse Spark 1.1
  21. 107.5 Claude Opus 4.1
  22. 107.5 Gemini 3.5 Flash
  23. 105.6 Kimi K3
  24. 103.9 Claude Opus 4.7
  25. 102.9 google/gemini-2.5-pro (snapshot unspecified)
  26. 102.8 Kimi K3.1 Pro
  27. 102.3 Muse Spark 1.2
  28. 102.2 Claude Opus 4.8
  29. 100.8 Muse Spark 1.3
  30. 100.8 Claude Sonnet 4.5
  31. 97.7 Muse Wonder
  32. 96.5 Claude Sonnet 4
  33. 95.7 Muse Spark Open
  34. 95.1 GPT-5.6 Terra
  35. 95.1 Claude Sonnet 5.5
  36. 94.1 GPT-5.4
  37. 94.0 Tencent Hy4
  38. 91.1 Claude Sonnet 5
  39. 90.0 Claude Sonnet 4.6
  40. 87.0 Qwen 3.6 Max (Preview)
  41. 86.8 Grok 4.6
  42. 86.5 Gemini 3.1 Flash-Lite
  43. 86.0 Qwen3.8-2.4T-A95B
  44. 85.8 GPT-5.2
  45. 85.8 Qwen3.7-Max
  46. 85.4 Qwen 3.8 Max
  47. 85.2 GPT-5.3 Codex
  48. 84.1 DeepSeek V4 Pro 0813
  49. 84.1 DeepSeek-V4-Pro
  50. 83.3 Qwen3-Max
  51. 83.1 Qwen3.7-Plus
  52. 82.8 Tencent Hy4 preview
  53. 82.5 LongCat-2.0
  54. 82.2 Qwen 3.6 Plus
  55. 81.7 Gemma 4 31B IT
  56. 81.6 DeepSeek V4.1 Flash 1017
  57. 81.3 DeepSeek V4.1 Flash
  58. 81.0 Grok 4.5
  59. 80.8 kwaipilot/kat-coder-pro-v2
  60. 80.6 Gemini 3.5 Flash-Lite
  61. 80.6 Qwen3.5-Plus
  62. 80.6 Kimi K2.5
  63. 80.0 Qwen3.5 397B-A17B
  64. 79.2 Kimi K2.6
  65. 79.1 DeepSeek-V4-Flash-Base
  66. 78.8 google/gemini-2.5-flash (snapshot unspecified)
  67. 78.3 nvidia/nemotron-3-ultra-550b-a55b
  68. 78.3 Inkling
  69. 78.1 DeepSeek-V4-Flash
  70. 78.0 moonshotai/kimi-k2 (snapshot unspecified)
  71. 77.6 DeepSeek-V3.1-Terminus
  72. 77.3 GLM-5.2
  73. 77.3 GPT-6 Luna
  74. 77.3 Qwen3-VL-235B-A22B-Instruct
  75. 77.1 MiMo-V2.5-Pro
  76. 76.7 DeepSeek-V3.2-Exp
  77. 76.6 GPT-5.6 Luna
  78. 76.5 deepseek/deepseek-chat-v3.1
  79. 76.4 GLM-5.3-Flash
  80. 76.2 GLM-5
  81. 75.5 DeepSeek V4 Flash 0731
  82. 75.2 GLM-5.1
  83. 75.1 GPT-4.1
  84. 75.0 DeepSeek-V3.2
  85. 74.6 Qwen3-235B-A22B-Instruct
  86. 74.5 GLM-5.3
  87. 74.4 Qwen3.8 Flash
  88. 74.4 deepseek/deepseek-chat-v3-0324
  89. 74.2 openai/gpt-4o (snapshot unspecified)
  90. 74.2 GPT-4o
  91. 74.2 Claude Haiku 4.5
  92. 73.9 Ling-2.6-1T
  93. 73.7 openai/gpt-4 (snapshot unspecified)
  94. 73.2 WizardLM-2 8x22B
  95. 73.1 Qwen3-Coder-480B-A35B
  96. 72.9 deepseek/deepseek-chat
  97. 72.8 GPT-4
  98. 72.6 MiniMax-M3
  99. 72.5 Tencent Hy3
  100. 72.3 Mercury 3
  101. 72.3 Tencent Hy3 preview
  102. 72.2 GPT-4o
  103. 71.3 GLM-4.7
  104. 71.3 openai/gpt-4-turbo (snapshot unspecified)
  105. 71.3 GLM-4.5
  106. 71.2 Qwen3-Coder-Plus
  107. 70.8 GLM-4.6
  108. 70.6 GPT-5.1
  109. 70.5 Qwen3-235B-A22B
  110. 70.5 Llama 4 Maverick
  111. 70.5 GPT-5.4 Mini
  112. 70.3 GPT-4o
  113. 70.2 Inkling-Small
  114. 70.2 DiffusionGemma 26B A4B
  115. 69.9 MiMo-V2.5
  116. 69.5 Gemma 4 26B A4B
  117. 69.4 Qwen3.5-122B-A10B
  118. 69.2 Grok 4.3
  119. 68.9 Qwen3-235B-A22B-Thinking
  120. 68.4 Qwen3-Next-80B-A3B-Instruct
  121. 68.4 MiniMax-M3.1-Flash
  122. 68.3 Claude Haiku 5.5
  123. 67.9 Muse Glimmer
  124. 67.5 Mistral Medium 3.5
  125. 67.4 Qwen3.6 27B
  126. 67.1 Grok 4.20
  127. 67.0 GPT-5
  128. 66.9 baidu/ernie-4.5-vl-424b-a47b
  129. 66.7 GLM-4.5V
  130. 66.6 Qwen3.5-27B
  131. 66.3 Ling-3.0-flash
  132. 66.2 nvidia/nemotron-3-super-120b-a12b
  133. 66.1 Qwen 3.8 27B
  134. 65.8 Mercury 2.5
  135. 65.5 inclusionai/ling-2.6-flash
  136. 65.5 Qwen3-Coder-Next
  137. 65.3 Solar Pro 4
  138. 65.2 Qwen2.5-VL-72B
  139. 65.0 GPT-4.1 mini
  140. 65.0 MiniMax-M2.5
  141. 64.8 reflection/beam-1.0-501b-a23b
  142. 64.6 nousresearch/hermes-3-llama-3.1-405b
  143. 64.5 Qwen3.5-35B-A3B-Base
  144. 64.5 google/gemini-2.5-flash-lite (snapshot unspecified)
  145. 64.0 MiniMax-M2.7
  146. 64.0 Qwen3-VL-32B-Instruct
  147. 63.8 Qwen3-Next-80B-A3B
  148. 63.8 sao10k/l3.3-euryale-70b
  149. 63.5 mistralai/mistral-small-2603
  150. 63.5 Qwen2.5-72B
  151. 63.4 GLM-4.6V
  152. 63.3 Qwen 3.6 Flash
  153. 63.3 Qwen3.7 Flash
  154. 62.8 Hermes 4 405B
  155. 62.8 meta-llama/llama-3.1-70b-instruct
  156. 62.7 local/qwen2.5-32b-instruct
  157. 62.6 Llama 3.3 70B Instruct
  158. 62.5 Laguna S 2.1
  159. 62.4 GLM-4.5-Air
  160. 62.1 Qwen3-30B-A3B-Instruct
  161. 61.6 Qwen 3.6 35B-A3B
  162. 61.4 Qwen3.5-Flash
  163. 61.3 amazon/nova-pro-v1
  164. 61.1 Qwen2.5-32B
  165. 61.0 sao10k/l3.1-euryale-70b
  166. 61.0 Qwen2-72B
  167. 60.1 Qwen3.5-35B-A3B
  168. 60.0 Qwen2.5-Coder-32B-Instruct
  169. 59.9 Qwen3-VL-30B-A3B-Instruct
  170. 59.7 Gemma 4 12B
  171. 59.6 Qwen3-Coder-30B-A3B-Instruct
  172. 59.5 Qwen3-Coder-Flash
  173. 59.3 Ripple-65B Preview
  174. 59.2 Qwen3-32B
  175. 59.1 thedrummer/cydonia-24b-v4.1
  176. 59.0 gpt-oss-120b
  177. 58.7 mistralai/ministral-14b-2512
  178. 58.3 Phi-4
  179. 57.9 nousresearch/hermes-4-70b
  180. 57.7 Llama 4 Scout
  181. 57.6 Llama 3.1-70B
  182. 57.5 Laguna XS 2.1
  183. 57.5 Gemma 3 27B PT
  184. 57.3 Claude 3 Haiku
  185. 56.2 local/mistral-small-24b-base-2501
  186. 56.1 Nemotron 3.5 Lightning 30B-A3B
  187. 56.0 Qwen3-VL-8B-Instruct
  188. 56.0 Gemma 3 27B
  189. 55.9 glm-4.7-flash
  190. 55.7 GPT-4o mini
  191. 55.4 mistralai/mistral-small-24b-instruct-2501
  192. 55.2 openai/gpt-3.5-turbo (snapshot unspecified)
  193. 55.1 Mercury 2
  194. 55.0 Qwen3-14B
  195. 54.8 gpt-oss-20b
  196. 54.3 Olmo 3.1 32B Instruct
  197. 54.2 Qwen3.5-9B-Base
  198. 54.0 local/olmo-3-32b-dpo
  199. 53.8 mistralai/mistral-small-3.2-24b-instruct
  200. 53.8 local/olmo-3-32b-rl
  201. 53.3 local/olmo-3-32b-mid
  202. 53.3 Gemma 2 27B
  203. 52.8 local/olmo-3-32b-sft
  204. 52.6 google/gemma-2-27b-it
  205. 52.6 local/nanbeige42-r2
  206. 52.6 local/qwen1.5-110b-chat
  207. 52.4 Qwen3.5-9B
  208. 51.7 local/ouro26-r4
  209. 51.4 Qwen3-8B
  210. 51.0 mistralai/ministral-8b-2512
  211. 50.7 GPT-5 mini
  212. 50.6 Gemma 4 E4B
  213. 50.4 Qwen3-30B-A3B
  214. 50.2 GPT-4.1 nano
  215. 49.9 Gemma 3 12B PT
  216. 49.7 Qwen1.5-72B
  217. 49.6 GPT-3.5 Turbo Instruct
  218. 49.4 GPT-5.4 Nano
  219. 49.3 Gemma 3 12B
  220. 48.7 Hunyuan-A13B-Instruct
  221. 48.4 qwen/qwen-2.5-7b-instruct
  222. 47.5 local/gemma-2-9b-it
  223. 47.3 Nemotron 3 Nano 30B-A3B
  224. 47.3 local/ouro14-r4
  225. 46.2 mistralai/ministral-3b-2512
  226. 46.0 Qwen3.5-4B-Base
  227. 45.0 local/olmo-3-32b-pretrain
  228. 44.7 Gemma 2 9B
  229. 44.3 Qwen3.5-4B
  230. 44.0 Gemma 4 E2B
  231. 43.9 UI-TARS-1.5
  232. 43.2 Mistral-Nemo-Base-2407
  233. 42.8 thedrummer/unslopnemo-12b
  234. 42.5 Mistral NeMo
  235. 41.8 local/llama-3.1-8b-instruct-fp8
  236. 41.6 meta-llama/llama-3.1-8b-instruct
  237. 40.9 local/llama-3.1-8b-instruct
  238. 40.8 local/olmo3-32b-step260000
  239. 40.7 local/llama-2-70b-hf
  240. 40.7 sao10k/l3-lunaris-8b
  241. 40.6 Llama 3.1-8B
  242. 39.2 local/llama-3-8b-instruct
  243. 38.5 Gemma 3n E4B
  244. 38.5 local/olmo3-32b-step360000
  245. 37.9 Gemma 3 4B
  246. 37.5 LLaMA-65B
  247. 37.2 aion-labs/aion-rp-llama-3.1-8b
  248. 37.2 ibm-granite/granite-4.0-h-micro
  249. 36.2 local/olmo3-32b-step130000
  250. 35.5 Qwen3.5-2B
  251. 34.5 local/olmo3-32b-step92000
  252. 34.4 local/olmo3-32b-step182000
  253. 32.6 davinci-002
  254. 32.0 local/olmo3-7b-step1413814
  255. 31.4 local/olmo3-32b-step65000
  256. 31.0 GPT-5 nano
  257. 29.5 local/olmo3-32b-step46000
  258. 27.4 local/olmo3-32b-step33000
  259. 25.8 local/gemma-7b-it
  260. 25.7 undi95/remm-slerp-l2-13b
  261. 24.9 Qwen3.5-0.8B
  262. 24.6 local/olmo3-7b-step360000
  263. 24.3 local/gemma-2-2b
  264. 24.1 Gemma 2 2B IT
  265. 23.9 local/talkie-1930-13b-it
  266. 23.8 gryphe/mythomax-l2-13b
  267. 23.2 local/olmo3-7b-step714000
  268. 22.7 local/olmo3-7b-step182000
  269. 21.5 local/olmo3-32b-step16000
  270. 20.4 local/olmo3-32b-step23000
  271. 18.9 local/olmo3-7b-step92000
  272. 18.7 local/llama-2-7b-hf
  273. 16.8 local/huginn-r12
  274. 16.1 Gemma 3 1B
  275. 15.6 local/olmo3-32b-step12000
  276. 15.0 LLaMA-7B
  277. 11.9 local/gemma-2b-it
  278. 10.6 local/talkie-web-13b-base
  279. 6.6 openai/babbage-002
  280. 4.1 local/olmo3-7b-step46000
  281. 3.1 local/olmo3-7b-step23000
  282. -8.3 local/olmo3-7b-step12000
  283. -8.6 local/pythia-12b
  284. -9.9 local/olmo3-32b-step6000
  285. -11.5 local/olmo3-7b-step6000
  286. -12.4 local/talkie-1930-13b-base
  287. -23.1 Gemma 3 270M
  1. 1 GPT-6 Astra NCRI 15.2; dots@2048; placed 173.9
  2. 2 GPT-6.1 Sol NCRI 15.2; dots@128; placed_after_release 156.6
  3. 3 Gemini 4 Argon NCRI 15.2; qrep@512; placed_after_release 154.1
  4. 4 GPT-6.1 Terra NCRI 15.2; qrep@512; placed_after_release 149.6
  5. 5 Gemini 3.8 Flash NCRI 15.2; dots@128; placed 129.7
  6. 6 Claude Fable 5.1 NCRI 15.2; dots@128; placed 129.6
  7. 7 Gemini 3.7 Flash NCRI 15.2; dots@32; sealed 127.6
  8. 8 Gemini 3.1 Pro NCRI 15.2; count@512; sealed 123.9
  9. 9 GPT-6 Sol NCRI 15.2; dots@128; placed_after_release 122.2
  10. 10 Claude Fable 5 NCRI 15.2; dots@32; sealed 121.5
  11. 11 GPT-5.6 Sol NCRI 15.2; qrep@512; sealed 119.5
  12. 12 Claude Opus 5.5 NCRI 15.2; dots@512; placed_after_release 118.9
  13. 13 Claude Opus 5 NCRI 15.2; dots@128; sealed 115.8
  14. 14 GPT-5.5 NCRI 15.2; qrep@512; sealed 111.0
  15. 15 Gemini 3.6 Flash NCRI 15.2; count@512; sealed 110.9
  16. 16 Claude Opus 4.6 NCRI 15.2; qrep@512; sealed 110.8
  17. 17 Gemini 3 Flash NCRI 15.2; count@512; sealed 110.5
  18. 18 Claude Opus 4.5 NCRI 15.2; dots@512; sealed 110.5
  19. 19 Claude Opus 4 NCRI 15.2; dots@512; sealed 109.0
  20. 20 Muse Spark 1.1 NCRI 15.2; qrep@512; sealed 108.8
  21. 21 Claude Opus 4.1 NCRI 15.2; dots@128; sealed 107.5
  22. 22 Gemini 3.5 Flash NCRI 15.2; dots@512; sealed 107.5
  23. 23 Kimi K3 NCRI 15.2; qrep@512; sealed 105.6
  24. 24 Claude Opus 4.7 NCRI 15.2; dots@128; sealed 103.9
  25. 25 google/gemini-2.5-pro (snapshot unspecified) NCRI 15.2; dots@512; sealed 102.9
  26. 26 Kimi K3.1 Pro NCRI 15.2; no filler; placed_after_release 102.8
  27. 27 Muse Spark 1.2 NCRI 15.2; dots@128; sealed 102.3
  28. 28 Claude Opus 4.8 NCRI 15.2; dots@128; sealed 102.2
  29. 29 Muse Spark 1.3 NCRI 15.2; dots@2048; placed 100.8
  30. 30 Claude Sonnet 4.5 NCRI 15.2; qrep@512; sealed 100.8
  31. 31 Muse Wonder NCRI 15.2; no filler; placed_after_release 97.7
  32. 32 Claude Sonnet 4 NCRI 15.2; count@512; sealed 96.5
  33. 33 Muse Spark Open NCRI 15.2; no filler; placed_after_release 95.7
  34. 34 GPT-5.6 Terra NCRI 15.2; qrep@512; sealed 95.1
  35. 35 Claude Sonnet 5.5 NCRI 15.2; no filler; placed_after_release 95.1
  36. 36 GPT-5.4 NCRI 15.2; dots@128; sealed 94.1
  37. 37 Tencent Hy4 NCRI 15.2; no filler; placed_after_release 94.0
  38. 38 Claude Sonnet 5 NCRI 15.2; dots@128; sealed 91.1
  39. 39 Claude Sonnet 4.6 NCRI 15.2; dots@512; sealed 90.0
  40. 40 Qwen 3.6 Max (Preview) NCRI 15.2; no filler; sealed 87.0
  41. 41 Grok 4.6 NCRI 15.2; no filler; sealed 86.8
  42. 42 Gemini 3.1 Flash-Lite NCRI 15.2; no filler; sealed 86.5
  43. 43 Qwen3.8-2.4T-A95B NCRI 15.2; no filler; sealed 86.0
  44. 44 GPT-5.2 NCRI 15.2; no filler; sealed 85.8
  45. 45 Qwen3.7-Max NCRI 15.2; no filler; sealed 85.8
  46. 46 Qwen 3.8 Max NCRI 15.2; no filler; sealed 85.4
  47. 47 GPT-5.3 Codex NCRI 15.2; no filler; sealed 85.2
  48. 48 DeepSeek V4 Pro 0813 NCRI 15.2; no filler; sealed 84.1
  49. 49 DeepSeek-V4-Pro NCRI 15.2; no filler; sealed 84.1
  50. 50 Qwen3-Max NCRI 15.2; no filler; sealed 83.3
  51. 51 Qwen3.7-Plus NCRI 15.2; no filler; sealed 83.1
  52. 52 Tencent Hy4 preview NCRI 15.2; no filler; sealed 82.8
  53. 53 LongCat-2.0 NCRI 15.2; no filler; sealed 82.5
  54. 54 Qwen 3.6 Plus NCRI 15.2; no filler; sealed 82.2
  55. 55 Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.7
  56. 56 DeepSeek V4.1 Flash 1017 NCRI 15.2; no filler; placed_after_release 81.6
  57. 57 DeepSeek V4.1 Flash NCRI 15.2; no filler; placed_after_release 81.3
  58. 58 Grok 4.5 NCRI 15.2; no filler; sealed 81.0
  59. 59 kwaipilot/kat-coder-pro-v2 NCRI 15.2; no filler; sealed 80.8
  60. 60 Gemini 3.5 Flash-Lite NCRI 15.2; no filler; sealed 80.6
  61. 61 Qwen3.5-Plus NCRI 15.2; no filler; sealed 80.6
  62. 62 Kimi K2.5 NCRI 15.2; no filler; sealed 80.6
  63. 63 Qwen3.5 397B-A17B NCRI 15.2; no filler; sealed 80.0
  64. 64 Kimi K2.6 NCRI 15.2; no filler; sealed 79.2
  65. 65 DeepSeek-V4-Flash-Base NCRI 15.2; no filler; sealed 79.1
  66. 66 google/gemini-2.5-flash (snapshot unspecified) NCRI 15.2; no filler; sealed 78.8
  67. 67 nvidia/nemotron-3-ultra-550b-a55b NCRI 15.2; no filler; sealed 78.3
  68. 68 Inkling NCRI 15.2; no filler; sealed 78.3
  69. 69 DeepSeek-V4-Flash NCRI 15.2; no filler; sealed 78.1
  70. 70 moonshotai/kimi-k2 (snapshot unspecified) NCRI 15.2; no filler; sealed 78.0
  71. 71 DeepSeek-V3.1-Terminus NCRI 15.2; no filler; sealed 77.6
  72. 72 GLM-5.2 NCRI 15.2; no filler; sealed 77.3
  73. 73 GPT-6 Luna NCRI 15.2; no filler; placed_after_release 77.3
  74. 74 Qwen3-VL-235B-A22B-Instruct NCRI 15.2; no filler; sealed 77.3
  75. 75 MiMo-V2.5-Pro NCRI 15.2; no filler; sealed 77.1
  76. 76 DeepSeek-V3.2-Exp NCRI 15.2; no filler; sealed 76.7
  77. 77 GPT-5.6 Luna NCRI 15.2; no filler; sealed 76.6
  78. 78 deepseek/deepseek-chat-v3.1 NCRI 15.2; no filler; sealed 76.5
  79. 79 GLM-5.3-Flash NCRI 15.2; no filler; sealed 76.4
  80. 80 GLM-5 NCRI 15.2; no filler; sealed 76.2
  81. 81 DeepSeek V4 Flash 0731 NCRI 15.2; no filler; sealed 75.5
  82. 82 GLM-5.1 NCRI 15.2; no filler; sealed 75.2
  83. 83 GPT-4.1 NCRI 15.2; no filler; sealed 75.1
  84. 84 DeepSeek-V3.2 NCRI 15.2; no filler; sealed 75.0
  85. 85 Qwen3-235B-A22B-Instruct Jul 2025, NCRI 15.2; no filler; sealed 74.6
  86. 86 GLM-5.3 NCRI 15.2; no filler; sealed 74.5
  87. 87 Qwen3.8 Flash NCRI 15.2; no filler; sealed 74.4
  88. 88 deepseek/deepseek-chat-v3-0324 NCRI 15.2; no filler; sealed 74.4
  89. 89 openai/gpt-4o (snapshot unspecified) NCRI 15.2; no filler; sealed 74.2
  90. 90 GPT-4o Aug 2024, NCRI 15.2; no filler; sealed 74.2
  91. 91 Claude Haiku 4.5 NCRI 15.2; no filler; sealed 74.2
  92. 92 Ling-2.6-1T NCRI 15.2; no filler; sealed 73.9
  93. 93 openai/gpt-4 (snapshot unspecified) NCRI 15.2; no filler; sealed 73.7
  94. 94 WizardLM-2 8x22B NCRI 15.2; no filler; sealed 73.2
  95. 95 Qwen3-Coder-480B-A35B NCRI 15.2; no filler; sealed 73.1
  96. 96 deepseek/deepseek-chat NCRI 15.2; no filler; sealed 72.9
  97. 97 GPT-4 Jun 2023, NCRI 15.2; no filler; sealed 72.8
  98. 98 MiniMax-M3 NCRI 15.2; no filler; sealed 72.6
  99. 99 Tencent Hy3 NCRI 15.2; no filler; sealed 72.5
  100. 100 Mercury 3 NCRI 15.2; no filler; placed_after_release 72.3
  101. 101 Tencent Hy3 preview NCRI 15.2; no filler; sealed 72.3
  102. 102 GPT-4o May 2024, NCRI 15.2; no filler; sealed 72.2
  103. 103 GLM-4.7 NCRI 15.2; no filler; sealed 71.3
  104. 104 openai/gpt-4-turbo (snapshot unspecified) NCRI 15.2; no filler; sealed 71.3
  105. 105 GLM-4.5 NCRI 15.2; no filler; sealed 71.3
  106. 106 Qwen3-Coder-Plus NCRI 15.2; no filler; sealed 71.2
  107. 107 GLM-4.6 NCRI 15.2; no filler; sealed 70.8
  108. 108 GPT-5.1 NCRI 15.2; no filler; sealed 70.6
  109. 109 Qwen3-235B-A22B NCRI 15.2; no filler; sealed 70.5
  110. 110 Llama 4 Maverick NCRI 15.2; no filler; sealed 70.5
  111. 111 GPT-5.4 Mini NCRI 15.2; no filler; sealed 70.5
  112. 112 GPT-4o Nov 2024, NCRI 15.2; no filler; sealed 70.3
  113. 113 Inkling-Small NCRI 15.2; no filler; sealed 70.2
  114. 114 DiffusionGemma 26B A4B NCRI 15.2; no filler; sealed 70.2
  115. 115 MiMo-V2.5 NCRI 15.2; no filler; sealed 69.9
  116. 116 Gemma 4 26B A4B NCRI 15.2; no filler; sealed 69.5
  117. 117 Qwen3.5-122B-A10B NCRI 15.2; no filler; sealed 69.4
  118. 118 Grok 4.3 NCRI 15.2; no filler; sealed 69.2
  119. 119 Qwen3-235B-A22B-Thinking Jul 2025, NCRI 15.2; no filler; sealed 68.9
  120. 120 Qwen3-Next-80B-A3B-Instruct NCRI 15.2; no filler; sealed 68.4
  121. 121 MiniMax-M3.1-Flash NCRI 15.2; no filler; placed_after_release 68.4
  122. 122 Claude Haiku 5.5 NCRI 15.2; no filler; placed_after_release 68.3
  123. 123 Muse Glimmer NCRI 15.2; no filler; sealed 67.9
  124. 124 Mistral Medium 3.5 NCRI 15.2; no filler; sealed 67.5
  125. 125 Qwen3.6 27B NCRI 15.2; no filler; sealed 67.4
  126. 126 Grok 4.20 NCRI 15.2; no filler; sealed 67.1
  127. 127 GPT-5 NCRI 15.2; no filler; sealed 67.0
  128. 128 baidu/ernie-4.5-vl-424b-a47b NCRI 15.2; no filler; sealed 66.9
  129. 129 GLM-4.5V NCRI 15.2; no filler; sealed 66.7
  130. 130 Qwen3.5-27B NCRI 15.2; no filler; sealed 66.6
  131. 131 Ling-3.0-flash NCRI 15.2; no filler; sealed 66.3
  132. 132 nvidia/nemotron-3-super-120b-a12b NCRI 15.2; no filler; sealed 66.2
  133. 133 Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
  134. 134 Mercury 2.5 NCRI 15.2; no filler; placed 65.8
  135. 135 inclusionai/ling-2.6-flash NCRI 15.2; no filler; sealed 65.5
  136. 136 Qwen3-Coder-Next NCRI 15.2; no filler; sealed 65.5
  137. 137 Solar Pro 4 NCRI 15.2; no filler; sealed 65.3
  138. 138 Qwen2.5-VL-72B NCRI 15.2; no filler; sealed 65.2
  139. 139 GPT-4.1 mini NCRI 15.2; no filler; sealed 65.0
  140. 140 MiniMax-M2.5 NCRI 15.2; no filler; sealed 65.0
  141. 141 reflection/beam-1.0-501b-a23b NCRI 15.2; no filler; placed_after_release 64.8
  142. 142 nousresearch/hermes-3-llama-3.1-405b NCRI 15.2; no filler; sealed 64.6
  143. 143 Qwen3.5-35B-A3B-Base NCRI 15.2; no filler; sealed 64.5
  144. 144 google/gemini-2.5-flash-lite (snapshot unspecified) NCRI 15.2; no filler; sealed 64.5
  145. 145 MiniMax-M2.7 NCRI 15.2; no filler; sealed 64.0
  146. 146 Qwen3-VL-32B-Instruct NCRI 15.2; no filler; sealed 64.0
  147. 147 Qwen3-Next-80B-A3B NCRI 15.2; no filler; sealed 63.8
  148. 148 sao10k/l3.3-euryale-70b NCRI 15.2; no filler; sealed 63.8
  149. 149 mistralai/mistral-small-2603 NCRI 15.2; no filler; sealed 63.5
  150. 150 Qwen2.5-72B NCRI 15.2; no filler; sealed 63.5
  151. 151 GLM-4.6V NCRI 15.2; no filler; sealed 63.4
  152. 152 Qwen 3.6 Flash NCRI 15.2; no filler; sealed 63.3
  153. 153 Qwen3.7 Flash NCRI 15.2; no filler; sealed 63.3
  154. 154 Hermes 4 405B NCRI 15.2; no filler; sealed 62.8
  155. 155 meta-llama/llama-3.1-70b-instruct NCRI 15.2; no filler; sealed 62.8
  156. 156 local/qwen2.5-32b-instruct NCRI 15.2; no filler; sealed 62.7
  157. 157 Llama 3.3 70B Instruct NCRI 15.2; no filler; sealed 62.6
  158. 158 Laguna S 2.1 NCRI 15.2; no filler; sealed 62.5
  159. 159 GLM-4.5-Air NCRI 15.2; no filler; sealed 62.4
  160. 160 Qwen3-30B-A3B-Instruct Jul 2025, NCRI 15.2; no filler; sealed 62.1
  161. 161 Qwen 3.6 35B-A3B NCRI 15.2; no filler; sealed 61.6
  162. 162 Qwen3.5-Flash NCRI 15.2; no filler; sealed 61.4
  163. 163 amazon/nova-pro-v1 NCRI 15.2; no filler; sealed 61.3
  164. 164 Qwen2.5-32B NCRI 15.2; no filler; sealed 61.1
  165. 165 sao10k/l3.1-euryale-70b NCRI 15.2; no filler; sealed 61.0
  166. 166 Qwen2-72B NCRI 15.2; no filler; sealed 61.0
  167. 167 Qwen3.5-35B-A3B NCRI 15.2; no filler; sealed 60.1
  168. 168 Qwen2.5-Coder-32B-Instruct NCRI 15.2; no filler; sealed 60.0
  169. 169 Qwen3-VL-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.9
  170. 170 Gemma 4 12B NCRI 15.2; no filler; sealed 59.7
  171. 171 Qwen3-Coder-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.6
  172. 172 Qwen3-Coder-Flash NCRI 15.2; no filler; sealed 59.5
  173. 173 Ripple-65B Preview NCRI 15.2; no filler; placed_after_release 59.3
  174. 174 Qwen3-32B NCRI 15.2; no filler; sealed 59.2
  175. 175 thedrummer/cydonia-24b-v4.1 NCRI 15.2; no filler; sealed 59.1
  176. 176 gpt-oss-120b NCRI 15.2; no filler; sealed 59.0
  177. 177 mistralai/ministral-14b-2512 NCRI 15.2; no filler; sealed 58.7
  178. 178 Phi-4 NCRI 15.2; no filler; sealed 58.3
  179. 179 nousresearch/hermes-4-70b NCRI 15.2; no filler; sealed 57.9
  180. 180 Llama 4 Scout NCRI 15.2; no filler; sealed 57.7
  181. 181 Llama 3.1-70B NCRI 15.2; no filler; sealed 57.6
  182. 182 Laguna XS 2.1 NCRI 15.2; no filler; sealed 57.5
  183. 183 Gemma 3 27B PT NCRI 15.2; no filler; sealed 57.5
  184. 184 Claude 3 Haiku NCRI 15.2; no filler; sealed 57.3
  185. 185 local/mistral-small-24b-base-2501 NCRI 15.2; no filler; sealed 56.2
  186. 186 Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 56.1
  187. 187 Qwen3-VL-8B-Instruct NCRI 15.2; no filler; sealed 56.0
  188. 188 Gemma 3 27B NCRI 15.2; no filler; sealed 56.0
  189. 189 glm-4.7-flash NCRI 15.2; no filler; sealed 55.9
  190. 190 GPT-4o mini NCRI 15.2; no filler; sealed 55.7
  191. 191 mistralai/mistral-small-24b-instruct-2501 NCRI 15.2; no filler; sealed 55.4
  192. 192 openai/gpt-3.5-turbo (snapshot unspecified) NCRI 15.2; no filler; sealed 55.2
  193. 193 Mercury 2 NCRI 15.2; no filler; sealed 55.1
  194. 194 Qwen3-14B NCRI 15.2; no filler; sealed 55.0
  195. 195 gpt-oss-20b NCRI 15.2; no filler; sealed 54.8
  196. 196 Olmo 3.1 32B Instruct NCRI 15.2; no filler; placed 54.3
  197. 197 Qwen3.5-9B-Base NCRI 15.2; no filler; sealed 54.2
  198. 198 local/olmo-3-32b-dpo NCRI 15.2; no filler; sealed 54.0
  199. 199 mistralai/mistral-small-3.2-24b-instruct NCRI 15.2; no filler; sealed 53.8
  200. 200 local/olmo-3-32b-rl NCRI 15.2; no filler; sealed 53.8
  201. 201 local/olmo-3-32b-mid NCRI 15.2; no filler; sealed 53.3
  202. 202 Gemma 2 27B NCRI 15.2; no filler; sealed 53.3
  203. 203 local/olmo-3-32b-sft NCRI 15.2; no filler; sealed 52.8
  204. 204 google/gemma-2-27b-it NCRI 15.2; no filler; sealed 52.6
  205. 205 local/nanbeige42-r2 NCRI 15.2; no filler; placed_after_release 52.6
  206. 206 local/qwen1.5-110b-chat NCRI 15.2; no filler; sealed 52.6
  207. 207 Qwen3.5-9B NCRI 15.2; no filler; sealed 52.4
  208. 208 local/ouro26-r4 NCRI 15.2; no filler; placed_after_release 51.7
  209. 209 Qwen3-8B NCRI 15.2; no filler; sealed 51.4
  210. 210 mistralai/ministral-8b-2512 NCRI 15.2; no filler; sealed 51.0
  211. 211 GPT-5 mini NCRI 15.2; no filler; sealed 50.7
  212. 212 Gemma 4 E4B NCRI 15.2; no filler; sealed 50.6
  213. 213 Qwen3-30B-A3B NCRI 15.2; no filler; sealed 50.4
  214. 214 GPT-4.1 nano NCRI 15.2; no filler; sealed 50.2
  215. 215 Gemma 3 12B PT NCRI 15.2; no filler; sealed 49.9
  216. 216 Qwen1.5-72B NCRI 15.2; no filler; sealed 49.7
  217. 217 GPT-3.5 Turbo Instruct NCRI 15.2; no filler; sealed 49.6
  218. 218 GPT-5.4 Nano NCRI 15.2; no filler; sealed 49.4
  219. 219 Gemma 3 12B NCRI 15.2; no filler; sealed 49.3
  220. 220 Hunyuan-A13B-Instruct NCRI 15.2; no filler; sealed 48.7
  221. 221 qwen/qwen-2.5-7b-instruct NCRI 15.2; no filler; sealed 48.4
  222. 222 local/gemma-2-9b-it NCRI 15.2; no filler; sealed 47.5
  223. 223 Nemotron 3 Nano 30B-A3B NCRI 15.2; no filler; sealed 47.3
  224. 224 local/ouro14-r4 NCRI 15.2; no filler; placed_after_release 47.3
  225. 225 mistralai/ministral-3b-2512 NCRI 15.2; no filler; sealed 46.2
  226. 226 Qwen3.5-4B-Base NCRI 15.2; no filler; sealed 46.0
  227. 227 local/olmo-3-32b-pretrain NCRI 15.2; no filler; sealed 45.0
  228. 228 Gemma 2 9B NCRI 15.2; no filler; sealed 44.7
  229. 229 Qwen3.5-4B NCRI 15.2; no filler; sealed 44.3
  230. 230 Gemma 4 E2B NCRI 15.2; no filler; sealed 44.0
  231. 231 UI-TARS-1.5 NCRI 15.2; no filler; sealed 43.9
  232. 232 Mistral-Nemo-Base-2407 NCRI 15.2; no filler; sealed 43.2
  233. 233 thedrummer/unslopnemo-12b NCRI 15.2; no filler; sealed 42.8
  234. 234 Mistral NeMo NCRI 15.2; no filler; sealed 42.5
  235. 235 local/llama-3.1-8b-instruct-fp8 NCRI 15.2; no filler; sealed 41.8
  236. 236 meta-llama/llama-3.1-8b-instruct NCRI 15.2; no filler; sealed 41.6
  237. 237 local/llama-3.1-8b-instruct NCRI 15.2; no filler; sealed 40.9
  238. 238 local/olmo3-32b-step260000 NCRI 15.2; no filler; placed 40.8
  239. 239 local/llama-2-70b-hf NCRI 15.2; no filler; sealed 40.7
  240. 240 sao10k/l3-lunaris-8b NCRI 15.2; no filler; sealed 40.7
  241. 241 Llama 3.1-8B NCRI 15.2; no filler; sealed 40.6
  242. 242 local/llama-3-8b-instruct NCRI 15.2; no filler; sealed 39.2
  243. 243 Gemma 3n E4B NCRI 15.2; no filler; sealed 38.5
  244. 244 local/olmo3-32b-step360000 NCRI 15.2; no filler; placed 38.5
  245. 245 Gemma 3 4B NCRI 15.2; no filler; sealed 37.9
  246. 246 LLaMA-65B NCRI 15.2; no filler; sealed 37.5
  247. 247 aion-labs/aion-rp-llama-3.1-8b NCRI 15.2; no filler; sealed 37.2
  248. 248 ibm-granite/granite-4.0-h-micro NCRI 15.2; no filler; sealed 37.2
  249. 249 local/olmo3-32b-step130000 NCRI 15.2; no filler; placed 36.2
  250. 250 Qwen3.5-2B NCRI 15.2; no filler; sealed 35.5
  251. 251 local/olmo3-32b-step92000 NCRI 15.2; no filler; placed 34.5
  252. 252 local/olmo3-32b-step182000 NCRI 15.2; no filler; sealed 34.4
  253. 253 davinci-002 NCRI 15.2; no filler; sealed 32.6
  254. 254 local/olmo3-7b-step1413814 NCRI 15.2; no filler; sealed 32.0
  255. 255 local/olmo3-32b-step65000 NCRI 15.2; no filler; placed 31.4
  256. 256 GPT-5 nano NCRI 15.2; no filler; sealed 31.0
  257. 257 local/olmo3-32b-step46000 NCRI 15.2; no filler; sealed 29.5
  258. 258 local/olmo3-32b-step33000 NCRI 15.2; no filler; placed 27.4
  259. 259 local/gemma-7b-it NCRI 15.2; no filler; sealed 25.8
  260. 260 undi95/remm-slerp-l2-13b NCRI 15.2; no filler; sealed 25.7
  261. 261 Qwen3.5-0.8B NCRI 15.2; no filler; sealed 24.9
  262. 262 local/olmo3-7b-step360000 NCRI 15.2; no filler; sealed 24.6
  263. 263 local/gemma-2-2b NCRI 15.2; no filler; sealed 24.3
  264. 264 Gemma 2 2B IT NCRI 15.2; no filler; sealed 24.1
  265. 265 local/talkie-1930-13b-it NCRI 15.2; no filler; sealed 23.9
  266. 266 gryphe/mythomax-l2-13b NCRI 15.2; no filler; sealed 23.8
  267. 267 local/olmo3-7b-step714000 NCRI 15.2; no filler; sealed 23.2
  268. 268 local/olmo3-7b-step182000 NCRI 15.2; no filler; sealed 22.7
  269. 269 local/olmo3-32b-step16000 NCRI 15.2; no filler; placed 21.5
  270. 270 local/olmo3-32b-step23000 NCRI 15.2; no filler; placed 20.4
  271. 271 local/olmo3-7b-step92000 NCRI 15.2; no filler; sealed 18.9
  272. 272 local/llama-2-7b-hf NCRI 15.2; no filler; sealed 18.7
  273. 273 local/huginn-r12 NCRI 15.2; no filler; placed_after_release 16.8
  274. 274 Gemma 3 1B NCRI 15.2; no filler; sealed 16.1
  275. 275 local/olmo3-32b-step12000 NCRI 15.2; no filler; placed 15.6
  276. 276 LLaMA-7B NCRI 15.2; no filler; sealed 15.0
  277. 277 local/gemma-2b-it NCRI 15.2; no filler; sealed 11.9
  278. 278 local/talkie-web-13b-base NCRI 15.2; no filler; sealed 10.6
  279. 279 openai/babbage-002 NCRI 15.2; no filler; sealed 6.6
  280. 280 local/olmo3-7b-step46000 NCRI 15.2; no filler; sealed 4.1
  281. 281 local/olmo3-7b-step23000 NCRI 15.2; no filler; sealed 3.1
  282. 282 local/olmo3-7b-step12000 NCRI 15.2; no filler; sealed -8.3
  283. 283 local/pythia-12b NCRI 15.2; no filler; sealed -8.6
  284. 284 local/olmo3-32b-step6000 NCRI 15.2; no filler; placed -9.9
  285. 285 local/olmo3-7b-step6000 NCRI 15.2; no filler; sealed -11.5
  286. 286 local/talkie-1930-13b-base NCRI 15.2; no filler; sealed -12.4
  287. 287 Gemma 3 270M NCRI 15.2; no filler; sealed -23.1

Source: Official nocot-bench (NCRI) leaderboard

201 models. Scroll sideways for the rest.

  1. 173.9 GPT-6 Astra
  2. 156.6 GPT-6.1 Sol
  3. 154.1 Gemini 4 Argon
  4. 149.6 GPT-6.1 Terra
  5. 129.7 Gemini 3.8 Flash
  6. 129.6 Claude Fable 5.1
  7. 127.6 Gemini 3.7 Flash
  8. 123.9 Gemini 3.1 Pro
  9. 122.2 GPT-6 Sol
  10. 121.5 Claude Fable 5
  11. 119.5 GPT-5.6 Sol
  12. 118.9 Claude Opus 5.5
  13. 115.8 Claude Opus 5
  14. 111.0 GPT-5.5
  15. 110.9 Gemini 3.6 Flash
  16. 110.8 Claude Opus 4.6
  17. 110.5 Gemini 3 Flash
  18. 110.5 Claude Opus 4.5
  19. 109.0 Claude Opus 4
  20. 108.8 Muse Spark 1.1
  21. 107.5 Claude Opus 4.1
  22. 107.5 Gemini 3.5 Flash
  23. 105.6 Kimi K3
  24. 103.9 Claude Opus 4.7
  25. 102.8 Kimi K3.1 Pro
  26. 102.3 Muse Spark 1.2
  27. 102.2 Claude Opus 4.8
  28. 100.8 Muse Spark 1.3
  29. 100.8 Claude Sonnet 4.5
  30. 97.7 Muse Wonder
  31. 96.5 Claude Sonnet 4
  32. 95.7 Muse Spark Open
  33. 95.1 GPT-5.6 Terra
  34. 95.1 Claude Sonnet 5.5
  35. 94.1 GPT-5.4
  36. 94.0 Tencent Hy4
  37. 91.1 Claude Sonnet 5
  38. 90.0 Claude Sonnet 4.6
  39. 87.0 Qwen 3.6 Max (Preview)
  40. 86.8 Grok 4.6
  41. 86.5 Gemini 3.1 Flash-Lite
  42. 86.0 Qwen3.8-2.4T-A95B
  43. 85.8 GPT-5.2
  44. 85.8 Qwen3.7-Max
  45. 85.4 Qwen 3.8 Max
  46. 85.2 GPT-5.3 Codex
  47. 84.1 DeepSeek V4 Pro 0813
  48. 84.1 DeepSeek-V4-Pro
  49. 83.3 Qwen3-Max
  50. 83.1 Qwen3.7-Plus
  51. 82.8 Tencent Hy4 preview
  52. 82.5 LongCat-2.0
  53. 82.2 Qwen 3.6 Plus
  54. 81.7 Gemma 4 31B IT
  55. 81.6 DeepSeek V4.1 Flash 1017
  56. 81.3 DeepSeek V4.1 Flash
  57. 81.0 Grok 4.5
  58. 80.6 Gemini 3.5 Flash-Lite
  59. 80.6 Qwen3.5-Plus
  60. 80.6 Kimi K2.5
  61. 80.0 Qwen3.5 397B-A17B
  62. 79.2 Kimi K2.6
  63. 79.1 DeepSeek-V4-Flash-Base
  64. 78.3 Inkling
  65. 78.1 DeepSeek-V4-Flash
  66. 77.6 DeepSeek-V3.1-Terminus
  67. 77.3 GLM-5.2
  68. 77.3 GPT-6 Luna
  69. 77.3 Qwen3-VL-235B-A22B-Instruct
  70. 77.1 MiMo-V2.5-Pro
  71. 76.7 DeepSeek-V3.2-Exp
  72. 76.6 GPT-5.6 Luna
  73. 76.4 GLM-5.3-Flash
  74. 76.2 GLM-5
  75. 75.5 DeepSeek V4 Flash 0731
  76. 75.2 GLM-5.1
  77. 75.1 GPT-4.1
  78. 75.0 DeepSeek-V3.2
  79. 74.6 Qwen3-235B-A22B-Instruct
  80. 74.5 GLM-5.3
  81. 74.4 Qwen3.8 Flash
  82. 74.2 GPT-4o
  83. 74.2 Claude Haiku 4.5
  84. 73.9 Ling-2.6-1T
  85. 73.2 WizardLM-2 8x22B
  86. 73.1 Qwen3-Coder-480B-A35B
  87. 72.8 GPT-4
  88. 72.6 MiniMax-M3
  89. 72.5 Tencent Hy3
  90. 72.3 Mercury 3
  91. 72.3 Tencent Hy3 preview
  92. 72.2 GPT-4o
  93. 71.3 GLM-4.7
  94. 71.3 GLM-4.5
  95. 71.2 Qwen3-Coder-Plus
  96. 70.8 GLM-4.6
  97. 70.6 GPT-5.1
  98. 70.5 Qwen3-235B-A22B
  99. 70.5 Llama 4 Maverick
  100. 70.5 GPT-5.4 Mini
  101. 70.3 GPT-4o
  102. 70.2 Inkling-Small
  103. 70.2 DiffusionGemma 26B A4B
  104. 69.9 MiMo-V2.5
  105. 69.5 Gemma 4 26B A4B
  106. 69.4 Qwen3.5-122B-A10B
  107. 69.2 Grok 4.3
  108. 68.9 Qwen3-235B-A22B-Thinking
  109. 68.4 Qwen3-Next-80B-A3B-Instruct
  110. 68.4 MiniMax-M3.1-Flash
  111. 68.3 Claude Haiku 5.5
  112. 67.9 Muse Glimmer
  113. 67.5 Mistral Medium 3.5
  114. 67.4 Qwen3.6 27B
  115. 67.1 Grok 4.20
  116. 67.0 GPT-5
  117. 66.7 GLM-4.5V
  118. 66.6 Qwen3.5-27B
  119. 66.3 Ling-3.0-flash
  120. 66.1 Qwen 3.8 27B
  121. 65.8 Mercury 2.5
  122. 65.5 Qwen3-Coder-Next
  123. 65.3 Solar Pro 4
  124. 65.2 Qwen2.5-VL-72B
  125. 65.0 GPT-4.1 mini
  126. 65.0 MiniMax-M2.5
  127. 64.5 Qwen3.5-35B-A3B-Base
  128. 64.0 MiniMax-M2.7
  129. 64.0 Qwen3-VL-32B-Instruct
  130. 63.8 Qwen3-Next-80B-A3B
  131. 63.5 Qwen2.5-72B
  132. 63.4 GLM-4.6V
  133. 63.3 Qwen 3.6 Flash
  134. 63.3 Qwen3.7 Flash
  135. 62.8 Hermes 4 405B
  136. 62.6 Llama 3.3 70B Instruct
  137. 62.5 Laguna S 2.1
  138. 62.4 GLM-4.5-Air
  139. 62.1 Qwen3-30B-A3B-Instruct
  140. 61.6 Qwen 3.6 35B-A3B
  141. 61.4 Qwen3.5-Flash
  142. 61.1 Qwen2.5-32B
  143. 61.0 Qwen2-72B
  144. 60.1 Qwen3.5-35B-A3B
  145. 60.0 Qwen2.5-Coder-32B-Instruct
  146. 59.9 Qwen3-VL-30B-A3B-Instruct
  147. 59.7 Gemma 4 12B
  148. 59.6 Qwen3-Coder-30B-A3B-Instruct
  149. 59.5 Qwen3-Coder-Flash
  150. 59.3 Ripple-65B Preview
  151. 59.2 Qwen3-32B
  152. 59.0 gpt-oss-120b
  153. 58.3 Phi-4
  154. 57.7 Llama 4 Scout
  155. 57.6 Llama 3.1-70B
  156. 57.5 Laguna XS 2.1
  157. 57.5 Gemma 3 27B PT
  158. 57.3 Claude 3 Haiku
  159. 56.1 Nemotron 3.5 Lightning 30B-A3B
  160. 56.0 Qwen3-VL-8B-Instruct
  161. 56.0 Gemma 3 27B
  162. 55.9 glm-4.7-flash
  163. 55.7 GPT-4o mini
  164. 55.1 Mercury 2
  165. 55.0 Qwen3-14B
  166. 54.8 gpt-oss-20b
  167. 54.3 Olmo 3.1 32B Instruct
  168. 54.2 Qwen3.5-9B-Base
  169. 53.3 Gemma 2 27B
  170. 52.4 Qwen3.5-9B
  171. 51.4 Qwen3-8B
  172. 50.7 GPT-5 mini
  173. 50.6 Gemma 4 E4B
  174. 50.4 Qwen3-30B-A3B
  175. 50.2 GPT-4.1 nano
  176. 49.9 Gemma 3 12B PT
  177. 49.7 Qwen1.5-72B
  178. 49.6 GPT-3.5 Turbo Instruct
  179. 49.4 GPT-5.4 Nano
  180. 49.3 Gemma 3 12B
  181. 48.7 Hunyuan-A13B-Instruct
  182. 47.3 Nemotron 3 Nano 30B-A3B
  183. 46.0 Qwen3.5-4B-Base
  184. 44.7 Gemma 2 9B
  185. 44.3 Qwen3.5-4B
  186. 44.0 Gemma 4 E2B
  187. 43.9 UI-TARS-1.5
  188. 43.2 Mistral-Nemo-Base-2407
  189. 42.5 Mistral NeMo
  190. 40.6 Llama 3.1-8B
  191. 38.5 Gemma 3n E4B
  192. 37.9 Gemma 3 4B
  193. 37.5 LLaMA-65B
  194. 35.5 Qwen3.5-2B
  195. 32.6 davinci-002
  196. 31.0 GPT-5 nano
  197. 24.9 Qwen3.5-0.8B
  198. 24.1 Gemma 2 2B IT
  199. 16.1 Gemma 3 1B
  200. 15.0 LLaMA-7B
  201. -23.1 Gemma 3 270M
  1. 1 GPT-6 Astra NCRI 15.2; dots@2048; placed 173.9
  2. 2 GPT-6.1 Sol NCRI 15.2; dots@128; placed_after_release 156.6
  3. 3 Gemini 4 Argon NCRI 15.2; qrep@512; placed_after_release 154.1
  4. 4 GPT-6.1 Terra NCRI 15.2; qrep@512; placed_after_release 149.6
  5. 5 Gemini 3.8 Flash NCRI 15.2; dots@128; placed 129.7
  6. 6 Claude Fable 5.1 NCRI 15.2; dots@128; placed 129.6
  7. 7 Gemini 3.7 Flash NCRI 15.2; dots@32; sealed 127.6
  8. 8 Gemini 3.1 Pro NCRI 15.2; count@512; sealed 123.9
  9. 9 GPT-6 Sol NCRI 15.2; dots@128; placed_after_release 122.2
  10. 10 Claude Fable 5 NCRI 15.2; dots@32; sealed 121.5
  11. 11 GPT-5.6 Sol NCRI 15.2; qrep@512; sealed 119.5
  12. 12 Claude Opus 5.5 NCRI 15.2; dots@512; placed_after_release 118.9
  13. 13 Claude Opus 5 NCRI 15.2; dots@128; sealed 115.8
  14. 14 GPT-5.5 NCRI 15.2; qrep@512; sealed 111.0
  15. 15 Gemini 3.6 Flash NCRI 15.2; count@512; sealed 110.9
  16. 16 Claude Opus 4.6 NCRI 15.2; qrep@512; sealed 110.8
  17. 17 Gemini 3 Flash NCRI 15.2; count@512; sealed 110.5
  18. 18 Claude Opus 4.5 NCRI 15.2; dots@512; sealed 110.5
  19. 19 Claude Opus 4 NCRI 15.2; dots@512; sealed 109.0
  20. 20 Muse Spark 1.1 NCRI 15.2; qrep@512; sealed 108.8
  21. 21 Claude Opus 4.1 NCRI 15.2; dots@128; sealed 107.5
  22. 22 Gemini 3.5 Flash NCRI 15.2; dots@512; sealed 107.5
  23. 23 Kimi K3 NCRI 15.2; qrep@512; sealed 105.6
  24. 24 Claude Opus 4.7 NCRI 15.2; dots@128; sealed 103.9
  25. 25 Kimi K3.1 Pro NCRI 15.2; no filler; placed_after_release 102.8
  26. 26 Muse Spark 1.2 NCRI 15.2; dots@128; sealed 102.3
  27. 27 Claude Opus 4.8 NCRI 15.2; dots@128; sealed 102.2
  28. 28 Muse Spark 1.3 NCRI 15.2; dots@2048; placed 100.8
  29. 29 Claude Sonnet 4.5 NCRI 15.2; qrep@512; sealed 100.8
  30. 30 Muse Wonder NCRI 15.2; no filler; placed_after_release 97.7
  31. 31 Claude Sonnet 4 NCRI 15.2; count@512; sealed 96.5
  32. 32 Muse Spark Open NCRI 15.2; no filler; placed_after_release 95.7
  33. 33 GPT-5.6 Terra NCRI 15.2; qrep@512; sealed 95.1
  34. 34 Claude Sonnet 5.5 NCRI 15.2; no filler; placed_after_release 95.1
  35. 35 GPT-5.4 NCRI 15.2; dots@128; sealed 94.1
  36. 36 Tencent Hy4 NCRI 15.2; no filler; placed_after_release 94.0
  37. 37 Claude Sonnet 5 NCRI 15.2; dots@128; sealed 91.1
  38. 38 Claude Sonnet 4.6 NCRI 15.2; dots@512; sealed 90.0
  39. 39 Qwen 3.6 Max (Preview) NCRI 15.2; no filler; sealed 87.0
  40. 40 Grok 4.6 NCRI 15.2; no filler; sealed 86.8
  41. 41 Gemini 3.1 Flash-Lite NCRI 15.2; no filler; sealed 86.5
  42. 42 Qwen3.8-2.4T-A95B NCRI 15.2; no filler; sealed 86.0
  43. 43 GPT-5.2 NCRI 15.2; no filler; sealed 85.8
  44. 44 Qwen3.7-Max NCRI 15.2; no filler; sealed 85.8
  45. 45 Qwen 3.8 Max NCRI 15.2; no filler; sealed 85.4
  46. 46 GPT-5.3 Codex NCRI 15.2; no filler; sealed 85.2
  47. 47 DeepSeek V4 Pro 0813 NCRI 15.2; no filler; sealed 84.1
  48. 48 DeepSeek-V4-Pro NCRI 15.2; no filler; sealed 84.1
  49. 49 Qwen3-Max NCRI 15.2; no filler; sealed 83.3
  50. 50 Qwen3.7-Plus NCRI 15.2; no filler; sealed 83.1
  51. 51 Tencent Hy4 preview NCRI 15.2; no filler; sealed 82.8
  52. 52 LongCat-2.0 NCRI 15.2; no filler; sealed 82.5
  53. 53 Qwen 3.6 Plus NCRI 15.2; no filler; sealed 82.2
  54. 54 Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.7
  55. 55 DeepSeek V4.1 Flash 1017 NCRI 15.2; no filler; placed_after_release 81.6
  56. 56 DeepSeek V4.1 Flash NCRI 15.2; no filler; placed_after_release 81.3
  57. 57 Grok 4.5 NCRI 15.2; no filler; sealed 81.0
  58. 58 Gemini 3.5 Flash-Lite NCRI 15.2; no filler; sealed 80.6
  59. 59 Qwen3.5-Plus NCRI 15.2; no filler; sealed 80.6
  60. 60 Kimi K2.5 NCRI 15.2; no filler; sealed 80.6
  61. 61 Qwen3.5 397B-A17B NCRI 15.2; no filler; sealed 80.0
  62. 62 Kimi K2.6 NCRI 15.2; no filler; sealed 79.2
  63. 63 DeepSeek-V4-Flash-Base NCRI 15.2; no filler; sealed 79.1
  64. 64 Inkling NCRI 15.2; no filler; sealed 78.3
  65. 65 DeepSeek-V4-Flash NCRI 15.2; no filler; sealed 78.1
  66. 66 DeepSeek-V3.1-Terminus NCRI 15.2; no filler; sealed 77.6
  67. 67 GLM-5.2 NCRI 15.2; no filler; sealed 77.3
  68. 68 GPT-6 Luna NCRI 15.2; no filler; placed_after_release 77.3
  69. 69 Qwen3-VL-235B-A22B-Instruct NCRI 15.2; no filler; sealed 77.3
  70. 70 MiMo-V2.5-Pro NCRI 15.2; no filler; sealed 77.1
  71. 71 DeepSeek-V3.2-Exp NCRI 15.2; no filler; sealed 76.7
  72. 72 GPT-5.6 Luna NCRI 15.2; no filler; sealed 76.6
  73. 73 GLM-5.3-Flash NCRI 15.2; no filler; sealed 76.4
  74. 74 GLM-5 NCRI 15.2; no filler; sealed 76.2
  75. 75 DeepSeek V4 Flash 0731 NCRI 15.2; no filler; sealed 75.5
  76. 76 GLM-5.1 NCRI 15.2; no filler; sealed 75.2
  77. 77 GPT-4.1 NCRI 15.2; no filler; sealed 75.1
  78. 78 DeepSeek-V3.2 NCRI 15.2; no filler; sealed 75.0
  79. 79 Qwen3-235B-A22B-Instruct Jul 2025, NCRI 15.2; no filler; sealed 74.6
  80. 80 GLM-5.3 NCRI 15.2; no filler; sealed 74.5
  81. 81 Qwen3.8 Flash NCRI 15.2; no filler; sealed 74.4
  82. 82 GPT-4o Aug 2024, NCRI 15.2; no filler; sealed 74.2
  83. 83 Claude Haiku 4.5 NCRI 15.2; no filler; sealed 74.2
  84. 84 Ling-2.6-1T NCRI 15.2; no filler; sealed 73.9
  85. 85 WizardLM-2 8x22B NCRI 15.2; no filler; sealed 73.2
  86. 86 Qwen3-Coder-480B-A35B NCRI 15.2; no filler; sealed 73.1
  87. 87 GPT-4 Jun 2023, NCRI 15.2; no filler; sealed 72.8
  88. 88 MiniMax-M3 NCRI 15.2; no filler; sealed 72.6
  89. 89 Tencent Hy3 NCRI 15.2; no filler; sealed 72.5
  90. 90 Mercury 3 NCRI 15.2; no filler; placed_after_release 72.3
  91. 91 Tencent Hy3 preview NCRI 15.2; no filler; sealed 72.3
  92. 92 GPT-4o May 2024, NCRI 15.2; no filler; sealed 72.2
  93. 93 GLM-4.7 NCRI 15.2; no filler; sealed 71.3
  94. 94 GLM-4.5 NCRI 15.2; no filler; sealed 71.3
  95. 95 Qwen3-Coder-Plus NCRI 15.2; no filler; sealed 71.2
  96. 96 GLM-4.6 NCRI 15.2; no filler; sealed 70.8
  97. 97 GPT-5.1 NCRI 15.2; no filler; sealed 70.6
  98. 98 Qwen3-235B-A22B NCRI 15.2; no filler; sealed 70.5
  99. 99 Llama 4 Maverick NCRI 15.2; no filler; sealed 70.5
  100. 100 GPT-5.4 Mini NCRI 15.2; no filler; sealed 70.5
  101. 101 GPT-4o Nov 2024, NCRI 15.2; no filler; sealed 70.3
  102. 102 Inkling-Small NCRI 15.2; no filler; sealed 70.2
  103. 103 DiffusionGemma 26B A4B NCRI 15.2; no filler; sealed 70.2
  104. 104 MiMo-V2.5 NCRI 15.2; no filler; sealed 69.9
  105. 105 Gemma 4 26B A4B NCRI 15.2; no filler; sealed 69.5
  106. 106 Qwen3.5-122B-A10B NCRI 15.2; no filler; sealed 69.4
  107. 107 Grok 4.3 NCRI 15.2; no filler; sealed 69.2
  108. 108 Qwen3-235B-A22B-Thinking Jul 2025, NCRI 15.2; no filler; sealed 68.9
  109. 109 Qwen3-Next-80B-A3B-Instruct NCRI 15.2; no filler; sealed 68.4
  110. 110 MiniMax-M3.1-Flash NCRI 15.2; no filler; placed_after_release 68.4
  111. 111 Claude Haiku 5.5 NCRI 15.2; no filler; placed_after_release 68.3
  112. 112 Muse Glimmer NCRI 15.2; no filler; sealed 67.9
  113. 113 Mistral Medium 3.5 NCRI 15.2; no filler; sealed 67.5
  114. 114 Qwen3.6 27B NCRI 15.2; no filler; sealed 67.4
  115. 115 Grok 4.20 NCRI 15.2; no filler; sealed 67.1
  116. 116 GPT-5 NCRI 15.2; no filler; sealed 67.0
  117. 117 GLM-4.5V NCRI 15.2; no filler; sealed 66.7
  118. 118 Qwen3.5-27B NCRI 15.2; no filler; sealed 66.6
  119. 119 Ling-3.0-flash NCRI 15.2; no filler; sealed 66.3
  120. 120 Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
  121. 121 Mercury 2.5 NCRI 15.2; no filler; placed 65.8
  122. 122 Qwen3-Coder-Next NCRI 15.2; no filler; sealed 65.5
  123. 123 Solar Pro 4 NCRI 15.2; no filler; sealed 65.3
  124. 124 Qwen2.5-VL-72B NCRI 15.2; no filler; sealed 65.2
  125. 125 GPT-4.1 mini NCRI 15.2; no filler; sealed 65.0
  126. 126 MiniMax-M2.5 NCRI 15.2; no filler; sealed 65.0
  127. 127 Qwen3.5-35B-A3B-Base NCRI 15.2; no filler; sealed 64.5
  128. 128 MiniMax-M2.7 NCRI 15.2; no filler; sealed 64.0
  129. 129 Qwen3-VL-32B-Instruct NCRI 15.2; no filler; sealed 64.0
  130. 130 Qwen3-Next-80B-A3B NCRI 15.2; no filler; sealed 63.8
  131. 131 Qwen2.5-72B NCRI 15.2; no filler; sealed 63.5
  132. 132 GLM-4.6V NCRI 15.2; no filler; sealed 63.4
  133. 133 Qwen 3.6 Flash NCRI 15.2; no filler; sealed 63.3
  134. 134 Qwen3.7 Flash NCRI 15.2; no filler; sealed 63.3
  135. 135 Hermes 4 405B NCRI 15.2; no filler; sealed 62.8
  136. 136 Llama 3.3 70B Instruct NCRI 15.2; no filler; sealed 62.6
  137. 137 Laguna S 2.1 NCRI 15.2; no filler; sealed 62.5
  138. 138 GLM-4.5-Air NCRI 15.2; no filler; sealed 62.4
  139. 139 Qwen3-30B-A3B-Instruct Jul 2025, NCRI 15.2; no filler; sealed 62.1
  140. 140 Qwen 3.6 35B-A3B NCRI 15.2; no filler; sealed 61.6
  141. 141 Qwen3.5-Flash NCRI 15.2; no filler; sealed 61.4
  142. 142 Qwen2.5-32B NCRI 15.2; no filler; sealed 61.1
  143. 143 Qwen2-72B NCRI 15.2; no filler; sealed 61.0
  144. 144 Qwen3.5-35B-A3B NCRI 15.2; no filler; sealed 60.1
  145. 145 Qwen2.5-Coder-32B-Instruct NCRI 15.2; no filler; sealed 60.0
  146. 146 Qwen3-VL-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.9
  147. 147 Gemma 4 12B NCRI 15.2; no filler; sealed 59.7
  148. 148 Qwen3-Coder-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.6
  149. 149 Qwen3-Coder-Flash NCRI 15.2; no filler; sealed 59.5
  150. 150 Ripple-65B Preview NCRI 15.2; no filler; placed_after_release 59.3
  151. 151 Qwen3-32B NCRI 15.2; no filler; sealed 59.2
  152. 152 gpt-oss-120b NCRI 15.2; no filler; sealed 59.0
  153. 153 Phi-4 NCRI 15.2; no filler; sealed 58.3
  154. 154 Llama 4 Scout NCRI 15.2; no filler; sealed 57.7
  155. 155 Llama 3.1-70B NCRI 15.2; no filler; sealed 57.6
  156. 156 Laguna XS 2.1 NCRI 15.2; no filler; sealed 57.5
  157. 157 Gemma 3 27B PT NCRI 15.2; no filler; sealed 57.5
  158. 158 Claude 3 Haiku NCRI 15.2; no filler; sealed 57.3
  159. 159 Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 56.1
  160. 160 Qwen3-VL-8B-Instruct NCRI 15.2; no filler; sealed 56.0
  161. 161 Gemma 3 27B NCRI 15.2; no filler; sealed 56.0
  162. 162 glm-4.7-flash NCRI 15.2; no filler; sealed 55.9
  163. 163 GPT-4o mini NCRI 15.2; no filler; sealed 55.7
  164. 164 Mercury 2 NCRI 15.2; no filler; sealed 55.1
  165. 165 Qwen3-14B NCRI 15.2; no filler; sealed 55.0
  166. 166 gpt-oss-20b NCRI 15.2; no filler; sealed 54.8
  167. 167 Olmo 3.1 32B Instruct NCRI 15.2; no filler; placed 54.3
  168. 168 Qwen3.5-9B-Base NCRI 15.2; no filler; sealed 54.2
  169. 169 Gemma 2 27B NCRI 15.2; no filler; sealed 53.3
  170. 170 Qwen3.5-9B NCRI 15.2; no filler; sealed 52.4
  171. 171 Qwen3-8B NCRI 15.2; no filler; sealed 51.4
  172. 172 GPT-5 mini NCRI 15.2; no filler; sealed 50.7
  173. 173 Gemma 4 E4B NCRI 15.2; no filler; sealed 50.6
  174. 174 Qwen3-30B-A3B NCRI 15.2; no filler; sealed 50.4
  175. 175 GPT-4.1 nano NCRI 15.2; no filler; sealed 50.2
  176. 176 Gemma 3 12B PT NCRI 15.2; no filler; sealed 49.9
  177. 177 Qwen1.5-72B NCRI 15.2; no filler; sealed 49.7
  178. 178 GPT-3.5 Turbo Instruct NCRI 15.2; no filler; sealed 49.6
  179. 179 GPT-5.4 Nano NCRI 15.2; no filler; sealed 49.4
  180. 180 Gemma 3 12B NCRI 15.2; no filler; sealed 49.3
  181. 181 Hunyuan-A13B-Instruct NCRI 15.2; no filler; sealed 48.7
  182. 182 Nemotron 3 Nano 30B-A3B NCRI 15.2; no filler; sealed 47.3
  183. 183 Qwen3.5-4B-Base NCRI 15.2; no filler; sealed 46.0
  184. 184 Gemma 2 9B NCRI 15.2; no filler; sealed 44.7
  185. 185 Qwen3.5-4B NCRI 15.2; no filler; sealed 44.3
  186. 186 Gemma 4 E2B NCRI 15.2; no filler; sealed 44.0
  187. 187 UI-TARS-1.5 NCRI 15.2; no filler; sealed 43.9
  188. 188 Mistral-Nemo-Base-2407 NCRI 15.2; no filler; sealed 43.2
  189. 189 Mistral NeMo NCRI 15.2; no filler; sealed 42.5
  190. 190 Llama 3.1-8B NCRI 15.2; no filler; sealed 40.6
  191. 191 Gemma 3n E4B NCRI 15.2; no filler; sealed 38.5
  192. 192 Gemma 3 4B NCRI 15.2; no filler; sealed 37.9
  193. 193 LLaMA-65B NCRI 15.2; no filler; sealed 37.5
  194. 194 Qwen3.5-2B NCRI 15.2; no filler; sealed 35.5
  195. 195 davinci-002 NCRI 15.2; no filler; sealed 32.6
  196. 196 GPT-5 nano NCRI 15.2; no filler; sealed 31.0
  197. 197 Qwen3.5-0.8B NCRI 15.2; no filler; sealed 24.9
  198. 198 Gemma 2 2B IT NCRI 15.2; no filler; sealed 24.1
  199. 199 Gemma 3 1B NCRI 15.2; no filler; sealed 16.1
  200. 200 LLaMA-7B NCRI 15.2; no filler; sealed 15.0
  201. 201 Gemma 3 270M NCRI 15.2; no filler; sealed -23.1

Source: Official nocot-bench (NCRI) leaderboard

202 models. Scroll sideways for the rest.

  1. 173.9 GPT-6 Astra
  2. 171.0 GPT-6.2 Astra
  3. 156.6 GPT-6.1 Sol
  4. 154.1 Gemini 4 Argon
  5. 149.6 GPT-6.1 Terra
  6. 129.7 Gemini 3.8 Flash
  7. 129.6 Claude Fable 5.1
  8. 127.6 Gemini 3.7 Flash
  9. 123.9 Gemini 3.1 Pro
  10. 122.2 GPT-6 Sol
  11. 121.5 Claude Fable 5
  12. 119.5 GPT-5.6 Sol
  13. 118.9 Claude Opus 5.5
  14. 115.8 Claude Opus 5
  15. 111.0 GPT-5.5
  16. 110.9 Gemini 3.6 Flash
  17. 110.8 Claude Opus 4.6
  18. 110.5 Gemini 3 Flash
  19. 110.5 Claude Opus 4.5
  20. 109.0 Claude Opus 4
  21. 108.8 Muse Spark 1.1
  22. 107.5 Claude Opus 4.1
  23. 107.5 Gemini 3.5 Flash
  24. 105.6 Kimi K3
  25. 103.9 Claude Opus 4.7
  26. 102.8 Kimi K3.1 Pro
  27. 102.3 Muse Spark 1.2
  28. 102.2 Claude Opus 4.8
  29. 100.8 Muse Spark 1.3
  30. 100.8 Claude Sonnet 4.5
  31. 97.7 Muse Wonder
  32. 96.5 Claude Sonnet 4
  33. 95.7 Muse Spark Open
  34. 95.1 GPT-5.6 Terra
  35. 95.1 Claude Sonnet 5.5
  36. 94.1 GPT-5.4
  37. 94.0 Tencent Hy4
  38. 91.1 Claude Sonnet 5
  39. 90.0 Claude Sonnet 4.6
  40. 87.0 Qwen 3.6 Max (Preview)
  41. 86.8 Grok 4.6
  42. 86.5 Gemini 3.1 Flash-Lite
  43. 86.0 Qwen3.8-2.4T-A95B
  44. 85.8 GPT-5.2
  45. 85.8 Qwen3.7-Max
  46. 85.4 Qwen 3.8 Max
  47. 85.2 GPT-5.3 Codex
  48. 84.1 DeepSeek V4 Pro 0813
  49. 84.1 DeepSeek-V4-Pro
  50. 83.3 Qwen3-Max
  51. 83.1 Qwen3.7-Plus
  52. 82.8 Tencent Hy4 preview
  53. 82.5 LongCat-2.0
  54. 82.2 Qwen 3.6 Plus
  55. 81.7 Gemma 4 31B IT
  56. 81.6 DeepSeek V4.1 Flash 1017
  57. 81.3 DeepSeek V4.1 Flash
  58. 81.0 Grok 4.5
  59. 80.6 Gemini 3.5 Flash-Lite
  60. 80.6 Qwen3.5-Plus
  61. 80.6 Kimi K2.5
  62. 80.0 Qwen3.5 397B-A17B
  63. 79.2 Kimi K2.6
  64. 79.1 DeepSeek-V4-Flash-Base
  65. 78.3 Inkling
  66. 78.1 DeepSeek-V4-Flash
  67. 77.6 DeepSeek-V3.1-Terminus
  68. 77.3 GLM-5.2
  69. 77.3 GPT-6 Luna
  70. 77.3 Qwen3-VL-235B-A22B-Instruct
  71. 77.1 MiMo-V2.5-Pro
  72. 76.7 DeepSeek-V3.2-Exp
  73. 76.6 GPT-5.6 Luna
  74. 76.4 GLM-5.3-Flash
  75. 76.2 GLM-5
  76. 75.5 DeepSeek V4 Flash 0731
  77. 75.2 GLM-5.1
  78. 75.1 GPT-4.1
  79. 75.0 DeepSeek-V3.2
  80. 74.6 Qwen3-235B-A22B-Instruct
  81. 74.5 GLM-5.3
  82. 74.4 Qwen3.8 Flash
  83. 74.2 GPT-4o
  84. 74.2 Claude Haiku 4.5
  85. 73.9 Ling-2.6-1T
  86. 73.2 WizardLM-2 8x22B
  87. 73.1 Qwen3-Coder-480B-A35B
  88. 72.8 GPT-4
  89. 72.6 MiniMax-M3
  90. 72.5 Tencent Hy3
  91. 72.3 Mercury 3
  92. 72.3 Tencent Hy3 preview
  93. 72.2 GPT-4o
  94. 71.3 GLM-4.7
  95. 71.3 GLM-4.5
  96. 71.2 Qwen3-Coder-Plus
  97. 70.8 GLM-4.6
  98. 70.6 GPT-5.1
  99. 70.5 Qwen3-235B-A22B
  100. 70.5 Llama 4 Maverick
  101. 70.5 GPT-5.4 Mini
  102. 70.3 GPT-4o
  103. 70.2 Inkling-Small
  104. 70.2 DiffusionGemma 26B A4B
  105. 69.9 MiMo-V2.5
  106. 69.5 Gemma 4 26B A4B
  107. 69.4 Qwen3.5-122B-A10B
  108. 69.2 Grok 4.3
  109. 68.9 Qwen3-235B-A22B-Thinking
  110. 68.4 Qwen3-Next-80B-A3B-Instruct
  111. 68.4 MiniMax-M3.1-Flash
  112. 68.3 Claude Haiku 5.5
  113. 67.9 Muse Glimmer
  114. 67.5 Mistral Medium 3.5
  115. 67.4 Qwen3.6 27B
  116. 67.1 Grok 4.20
  117. 67.0 GPT-5
  118. 66.7 GLM-4.5V
  119. 66.6 Qwen3.5-27B
  120. 66.3 Ling-3.0-flash
  121. 66.1 Qwen 3.8 27B
  122. 65.8 Mercury 2.5
  123. 65.5 Qwen3-Coder-Next
  124. 65.3 Solar Pro 4
  125. 65.2 Qwen2.5-VL-72B
  126. 65.0 GPT-4.1 mini
  127. 65.0 MiniMax-M2.5
  128. 64.5 Qwen3.5-35B-A3B-Base
  129. 64.0 MiniMax-M2.7
  130. 64.0 Qwen3-VL-32B-Instruct
  131. 63.8 Qwen3-Next-80B-A3B
  132. 63.5 Qwen2.5-72B
  133. 63.4 GLM-4.6V
  134. 63.3 Qwen 3.6 Flash
  135. 63.3 Qwen3.7 Flash
  136. 62.8 Hermes 4 405B
  137. 62.6 Llama 3.3 70B Instruct
  138. 62.5 Laguna S 2.1
  139. 62.4 GLM-4.5-Air
  140. 62.1 Qwen3-30B-A3B-Instruct
  141. 61.6 Qwen 3.6 35B-A3B
  142. 61.4 Qwen3.5-Flash
  143. 61.1 Qwen2.5-32B
  144. 61.0 Qwen2-72B
  145. 60.1 Qwen3.5-35B-A3B
  146. 60.0 Qwen2.5-Coder-32B-Instruct
  147. 59.9 Qwen3-VL-30B-A3B-Instruct
  148. 59.7 Gemma 4 12B
  149. 59.6 Qwen3-Coder-30B-A3B-Instruct
  150. 59.5 Qwen3-Coder-Flash
  151. 59.3 Ripple-65B Preview
  152. 59.2 Qwen3-32B
  153. 59.0 gpt-oss-120b
  154. 58.3 Phi-4
  155. 57.7 Llama 4 Scout
  156. 57.6 Llama 3.1-70B
  157. 57.5 Laguna XS 2.1
  158. 57.5 Gemma 3 27B PT
  159. 57.3 Claude 3 Haiku
  160. 56.1 Nemotron 3.5 Lightning 30B-A3B
  161. 56.0 Qwen3-VL-8B-Instruct
  162. 56.0 Gemma 3 27B
  163. 55.9 glm-4.7-flash
  164. 55.7 GPT-4o mini
  165. 55.1 Mercury 2
  166. 55.0 Qwen3-14B
  167. 54.8 gpt-oss-20b
  168. 54.3 Olmo 3.1 32B Instruct
  169. 54.2 Qwen3.5-9B-Base
  170. 53.3 Gemma 2 27B
  171. 52.4 Qwen3.5-9B
  172. 51.4 Qwen3-8B
  173. 50.7 GPT-5 mini
  174. 50.6 Gemma 4 E4B
  175. 50.4 Qwen3-30B-A3B
  176. 50.2 GPT-4.1 nano
  177. 49.9 Gemma 3 12B PT
  178. 49.7 Qwen1.5-72B
  179. 49.6 GPT-3.5 Turbo Instruct
  180. 49.4 GPT-5.4 Nano
  181. 49.3 Gemma 3 12B
  182. 48.7 Hunyuan-A13B-Instruct
  183. 47.3 Nemotron 3 Nano 30B-A3B
  184. 46.0 Qwen3.5-4B-Base
  185. 44.7 Gemma 2 9B
  186. 44.3 Qwen3.5-4B
  187. 44.0 Gemma 4 E2B
  188. 43.9 UI-TARS-1.5
  189. 43.2 Mistral-Nemo-Base-2407
  190. 42.5 Mistral NeMo
  191. 40.6 Llama 3.1-8B
  192. 38.5 Gemma 3n E4B
  193. 37.9 Gemma 3 4B
  194. 37.5 LLaMA-65B
  195. 35.5 Qwen3.5-2B
  196. 32.6 davinci-002
  197. 31.0 GPT-5 nano
  198. 24.9 Qwen3.5-0.8B
  199. 24.1 Gemma 2 2B IT
  200. 16.1 Gemma 3 1B
  201. 15.0 LLaMA-7B
  202. -23.1 Gemma 3 270M
  1. 1 GPT-6 Astra NCRI 15.2; dots@2048; placed 173.9
  2. 2 GPT-6.2 Astra 171.0
  3. 3 GPT-6.1 Sol NCRI 15.2; dots@128; placed_after_release 156.6
  4. 4 Gemini 4 Argon NCRI 15.2; qrep@512; placed_after_release 154.1
  5. 5 GPT-6.1 Terra NCRI 15.2; qrep@512; placed_after_release 149.6
  6. 6 Gemini 3.8 Flash NCRI 15.2; dots@128; placed 129.7
  7. 7 Claude Fable 5.1 NCRI 15.2; dots@128; placed 129.6
  8. 8 Gemini 3.7 Flash NCRI 15.2; dots@32; sealed 127.6
  9. 9 Gemini 3.1 Pro NCRI 15.2; count@512; sealed 123.9
  10. 10 GPT-6 Sol NCRI 15.2; dots@128; placed_after_release 122.2
  11. 11 Claude Fable 5 NCRI 15.2; dots@32; sealed 121.5
  12. 12 GPT-5.6 Sol NCRI 15.2; qrep@512; sealed 119.5
  13. 13 Claude Opus 5.5 NCRI 15.2; dots@512; placed_after_release 118.9
  14. 14 Claude Opus 5 NCRI 15.2; dots@128; sealed 115.8
  15. 15 GPT-5.5 NCRI 15.2; qrep@512; sealed 111.0
  16. 16 Gemini 3.6 Flash NCRI 15.2; count@512; sealed 110.9
  17. 17 Claude Opus 4.6 NCRI 15.2; qrep@512; sealed 110.8
  18. 18 Gemini 3 Flash NCRI 15.2; count@512; sealed 110.5
  19. 19 Claude Opus 4.5 NCRI 15.2; dots@512; sealed 110.5
  20. 20 Claude Opus 4 NCRI 15.2; dots@512; sealed 109.0
  21. 21 Muse Spark 1.1 NCRI 15.2; qrep@512; sealed 108.8
  22. 22 Claude Opus 4.1 NCRI 15.2; dots@128; sealed 107.5
  23. 23 Gemini 3.5 Flash NCRI 15.2; dots@512; sealed 107.5
  24. 24 Kimi K3 NCRI 15.2; qrep@512; sealed 105.6
  25. 25 Claude Opus 4.7 NCRI 15.2; dots@128; sealed 103.9
  26. 26 Kimi K3.1 Pro NCRI 15.2; no filler; placed_after_release 102.8
  27. 27 Muse Spark 1.2 NCRI 15.2; dots@128; sealed 102.3
  28. 28 Claude Opus 4.8 NCRI 15.2; dots@128; sealed 102.2
  29. 29 Muse Spark 1.3 NCRI 15.2; dots@2048; placed 100.8
  30. 30 Claude Sonnet 4.5 NCRI 15.2; qrep@512; sealed 100.8
  31. 31 Muse Wonder NCRI 15.2; no filler; placed_after_release 97.7
  32. 32 Claude Sonnet 4 NCRI 15.2; count@512; sealed 96.5
  33. 33 Muse Spark Open NCRI 15.2; no filler; placed_after_release 95.7
  34. 34 GPT-5.6 Terra NCRI 15.2; qrep@512; sealed 95.1
  35. 35 Claude Sonnet 5.5 NCRI 15.2; no filler; placed_after_release 95.1
  36. 36 GPT-5.4 NCRI 15.2; dots@128; sealed 94.1
  37. 37 Tencent Hy4 NCRI 15.2; no filler; placed_after_release 94.0
  38. 38 Claude Sonnet 5 NCRI 15.2; dots@128; sealed 91.1
  39. 39 Claude Sonnet 4.6 NCRI 15.2; dots@512; sealed 90.0
  40. 40 Qwen 3.6 Max (Preview) NCRI 15.2; no filler; sealed 87.0
  41. 41 Grok 4.6 NCRI 15.2; no filler; sealed 86.8
  42. 42 Gemini 3.1 Flash-Lite NCRI 15.2; no filler; sealed 86.5
  43. 43 Qwen3.8-2.4T-A95B NCRI 15.2; no filler; sealed 86.0
  44. 44 GPT-5.2 NCRI 15.2; no filler; sealed 85.8
  45. 45 Qwen3.7-Max NCRI 15.2; no filler; sealed 85.8
  46. 46 Qwen 3.8 Max NCRI 15.2; no filler; sealed 85.4
  47. 47 GPT-5.3 Codex NCRI 15.2; no filler; sealed 85.2
  48. 48 DeepSeek V4 Pro 0813 NCRI 15.2; no filler; sealed 84.1
  49. 49 DeepSeek-V4-Pro NCRI 15.2; no filler; sealed 84.1
  50. 50 Qwen3-Max NCRI 15.2; no filler; sealed 83.3
  51. 51 Qwen3.7-Plus NCRI 15.2; no filler; sealed 83.1
  52. 52 Tencent Hy4 preview NCRI 15.2; no filler; sealed 82.8
  53. 53 LongCat-2.0 NCRI 15.2; no filler; sealed 82.5
  54. 54 Qwen 3.6 Plus NCRI 15.2; no filler; sealed 82.2
  55. 55 Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.7
  56. 56 DeepSeek V4.1 Flash 1017 NCRI 15.2; no filler; placed_after_release 81.6
  57. 57 DeepSeek V4.1 Flash NCRI 15.2; no filler; placed_after_release 81.3
  58. 58 Grok 4.5 NCRI 15.2; no filler; sealed 81.0
  59. 59 Gemini 3.5 Flash-Lite NCRI 15.2; no filler; sealed 80.6
  60. 60 Qwen3.5-Plus NCRI 15.2; no filler; sealed 80.6
  61. 61 Kimi K2.5 NCRI 15.2; no filler; sealed 80.6
  62. 62 Qwen3.5 397B-A17B NCRI 15.2; no filler; sealed 80.0
  63. 63 Kimi K2.6 NCRI 15.2; no filler; sealed 79.2
  64. 64 DeepSeek-V4-Flash-Base NCRI 15.2; no filler; sealed 79.1
  65. 65 Inkling NCRI 15.2; no filler; sealed 78.3
  66. 66 DeepSeek-V4-Flash NCRI 15.2; no filler; sealed 78.1
  67. 67 DeepSeek-V3.1-Terminus NCRI 15.2; no filler; sealed 77.6
  68. 68 GLM-5.2 NCRI 15.2; no filler; sealed 77.3
  69. 69 GPT-6 Luna NCRI 15.2; no filler; placed_after_release 77.3
  70. 70 Qwen3-VL-235B-A22B-Instruct NCRI 15.2; no filler; sealed 77.3
  71. 71 MiMo-V2.5-Pro NCRI 15.2; no filler; sealed 77.1
  72. 72 DeepSeek-V3.2-Exp NCRI 15.2; no filler; sealed 76.7
  73. 73 GPT-5.6 Luna NCRI 15.2; no filler; sealed 76.6
  74. 74 GLM-5.3-Flash NCRI 15.2; no filler; sealed 76.4
  75. 75 GLM-5 NCRI 15.2; no filler; sealed 76.2
  76. 76 DeepSeek V4 Flash 0731 NCRI 15.2; no filler; sealed 75.5
  77. 77 GLM-5.1 NCRI 15.2; no filler; sealed 75.2
  78. 78 GPT-4.1 NCRI 15.2; no filler; sealed 75.1
  79. 79 DeepSeek-V3.2 NCRI 15.2; no filler; sealed 75.0
  80. 80 Qwen3-235B-A22B-Instruct Jul 2025, NCRI 15.2; no filler; sealed 74.6
  81. 81 GLM-5.3 NCRI 15.2; no filler; sealed 74.5
  82. 82 Qwen3.8 Flash NCRI 15.2; no filler; sealed 74.4
  83. 83 GPT-4o Aug 2024, NCRI 15.2; no filler; sealed 74.2
  84. 84 Claude Haiku 4.5 NCRI 15.2; no filler; sealed 74.2
  85. 85 Ling-2.6-1T NCRI 15.2; no filler; sealed 73.9
  86. 86 WizardLM-2 8x22B NCRI 15.2; no filler; sealed 73.2
  87. 87 Qwen3-Coder-480B-A35B NCRI 15.2; no filler; sealed 73.1
  88. 88 GPT-4 Jun 2023, NCRI 15.2; no filler; sealed 72.8
  89. 89 MiniMax-M3 NCRI 15.2; no filler; sealed 72.6
  90. 90 Tencent Hy3 NCRI 15.2; no filler; sealed 72.5
  91. 91 Mercury 3 NCRI 15.2; no filler; placed_after_release 72.3
  92. 92 Tencent Hy3 preview NCRI 15.2; no filler; sealed 72.3
  93. 93 GPT-4o May 2024, NCRI 15.2; no filler; sealed 72.2
  94. 94 GLM-4.7 NCRI 15.2; no filler; sealed 71.3
  95. 95 GLM-4.5 NCRI 15.2; no filler; sealed 71.3
  96. 96 Qwen3-Coder-Plus NCRI 15.2; no filler; sealed 71.2
  97. 97 GLM-4.6 NCRI 15.2; no filler; sealed 70.8
  98. 98 GPT-5.1 NCRI 15.2; no filler; sealed 70.6
  99. 99 Qwen3-235B-A22B NCRI 15.2; no filler; sealed 70.5
  100. 100 Llama 4 Maverick NCRI 15.2; no filler; sealed 70.5
  101. 101 GPT-5.4 Mini NCRI 15.2; no filler; sealed 70.5
  102. 102 GPT-4o Nov 2024, NCRI 15.2; no filler; sealed 70.3
  103. 103 Inkling-Small NCRI 15.2; no filler; sealed 70.2
  104. 104 DiffusionGemma 26B A4B NCRI 15.2; no filler; sealed 70.2
  105. 105 MiMo-V2.5 NCRI 15.2; no filler; sealed 69.9
  106. 106 Gemma 4 26B A4B NCRI 15.2; no filler; sealed 69.5
  107. 107 Qwen3.5-122B-A10B NCRI 15.2; no filler; sealed 69.4
  108. 108 Grok 4.3 NCRI 15.2; no filler; sealed 69.2
  109. 109 Qwen3-235B-A22B-Thinking Jul 2025, NCRI 15.2; no filler; sealed 68.9
  110. 110 Qwen3-Next-80B-A3B-Instruct NCRI 15.2; no filler; sealed 68.4
  111. 111 MiniMax-M3.1-Flash NCRI 15.2; no filler; placed_after_release 68.4
  112. 112 Claude Haiku 5.5 NCRI 15.2; no filler; placed_after_release 68.3
  113. 113 Muse Glimmer NCRI 15.2; no filler; sealed 67.9
  114. 114 Mistral Medium 3.5 NCRI 15.2; no filler; sealed 67.5
  115. 115 Qwen3.6 27B NCRI 15.2; no filler; sealed 67.4
  116. 116 Grok 4.20 NCRI 15.2; no filler; sealed 67.1
  117. 117 GPT-5 NCRI 15.2; no filler; sealed 67.0
  118. 118 GLM-4.5V NCRI 15.2; no filler; sealed 66.7
  119. 119 Qwen3.5-27B NCRI 15.2; no filler; sealed 66.6
  120. 120 Ling-3.0-flash NCRI 15.2; no filler; sealed 66.3
  121. 121 Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
  122. 122 Mercury 2.5 NCRI 15.2; no filler; placed 65.8
  123. 123 Qwen3-Coder-Next NCRI 15.2; no filler; sealed 65.5
  124. 124 Solar Pro 4 NCRI 15.2; no filler; sealed 65.3
  125. 125 Qwen2.5-VL-72B NCRI 15.2; no filler; sealed 65.2
  126. 126 GPT-4.1 mini NCRI 15.2; no filler; sealed 65.0
  127. 127 MiniMax-M2.5 NCRI 15.2; no filler; sealed 65.0
  128. 128 Qwen3.5-35B-A3B-Base NCRI 15.2; no filler; sealed 64.5
  129. 129 MiniMax-M2.7 NCRI 15.2; no filler; sealed 64.0
  130. 130 Qwen3-VL-32B-Instruct NCRI 15.2; no filler; sealed 64.0
  131. 131 Qwen3-Next-80B-A3B NCRI 15.2; no filler; sealed 63.8
  132. 132 Qwen2.5-72B NCRI 15.2; no filler; sealed 63.5
  133. 133 GLM-4.6V NCRI 15.2; no filler; sealed 63.4
  134. 134 Qwen 3.6 Flash NCRI 15.2; no filler; sealed 63.3
  135. 135 Qwen3.7 Flash NCRI 15.2; no filler; sealed 63.3
  136. 136 Hermes 4 405B NCRI 15.2; no filler; sealed 62.8
  137. 137 Llama 3.3 70B Instruct NCRI 15.2; no filler; sealed 62.6
  138. 138 Laguna S 2.1 NCRI 15.2; no filler; sealed 62.5
  139. 139 GLM-4.5-Air NCRI 15.2; no filler; sealed 62.4
  140. 140 Qwen3-30B-A3B-Instruct Jul 2025, NCRI 15.2; no filler; sealed 62.1
  141. 141 Qwen 3.6 35B-A3B NCRI 15.2; no filler; sealed 61.6
  142. 142 Qwen3.5-Flash NCRI 15.2; no filler; sealed 61.4
  143. 143 Qwen2.5-32B NCRI 15.2; no filler; sealed 61.1
  144. 144 Qwen2-72B NCRI 15.2; no filler; sealed 61.0
  145. 145 Qwen3.5-35B-A3B NCRI 15.2; no filler; sealed 60.1
  146. 146 Qwen2.5-Coder-32B-Instruct NCRI 15.2; no filler; sealed 60.0
  147. 147 Qwen3-VL-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.9
  148. 148 Gemma 4 12B NCRI 15.2; no filler; sealed 59.7
  149. 149 Qwen3-Coder-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.6
  150. 150 Qwen3-Coder-Flash NCRI 15.2; no filler; sealed 59.5
  151. 151 Ripple-65B Preview NCRI 15.2; no filler; placed_after_release 59.3
  152. 152 Qwen3-32B NCRI 15.2; no filler; sealed 59.2
  153. 153 gpt-oss-120b NCRI 15.2; no filler; sealed 59.0
  154. 154 Phi-4 NCRI 15.2; no filler; sealed 58.3
  155. 155 Llama 4 Scout NCRI 15.2; no filler; sealed 57.7
  156. 156 Llama 3.1-70B NCRI 15.2; no filler; sealed 57.6
  157. 157 Laguna XS 2.1 NCRI 15.2; no filler; sealed 57.5
  158. 158 Gemma 3 27B PT NCRI 15.2; no filler; sealed 57.5
  159. 159 Claude 3 Haiku NCRI 15.2; no filler; sealed 57.3
  160. 160 Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 56.1
  161. 161 Qwen3-VL-8B-Instruct NCRI 15.2; no filler; sealed 56.0
  162. 162 Gemma 3 27B NCRI 15.2; no filler; sealed 56.0
  163. 163 glm-4.7-flash NCRI 15.2; no filler; sealed 55.9
  164. 164 GPT-4o mini NCRI 15.2; no filler; sealed 55.7
  165. 165 Mercury 2 NCRI 15.2; no filler; sealed 55.1
  166. 166 Qwen3-14B NCRI 15.2; no filler; sealed 55.0
  167. 167 gpt-oss-20b NCRI 15.2; no filler; sealed 54.8
  168. 168 Olmo 3.1 32B Instruct NCRI 15.2; no filler; placed 54.3
  169. 169 Qwen3.5-9B-Base NCRI 15.2; no filler; sealed 54.2
  170. 170 Gemma 2 27B NCRI 15.2; no filler; sealed 53.3
  171. 171 Qwen3.5-9B NCRI 15.2; no filler; sealed 52.4
  172. 172 Qwen3-8B NCRI 15.2; no filler; sealed 51.4
  173. 173 GPT-5 mini NCRI 15.2; no filler; sealed 50.7
  174. 174 Gemma 4 E4B NCRI 15.2; no filler; sealed 50.6
  175. 175 Qwen3-30B-A3B NCRI 15.2; no filler; sealed 50.4
  176. 176 GPT-4.1 nano NCRI 15.2; no filler; sealed 50.2
  177. 177 Gemma 3 12B PT NCRI 15.2; no filler; sealed 49.9
  178. 178 Qwen1.5-72B NCRI 15.2; no filler; sealed 49.7
  179. 179 GPT-3.5 Turbo Instruct NCRI 15.2; no filler; sealed 49.6
  180. 180 GPT-5.4 Nano NCRI 15.2; no filler; sealed 49.4
  181. 181 Gemma 3 12B NCRI 15.2; no filler; sealed 49.3
  182. 182 Hunyuan-A13B-Instruct NCRI 15.2; no filler; sealed 48.7
  183. 183 Nemotron 3 Nano 30B-A3B NCRI 15.2; no filler; sealed 47.3
  184. 184 Qwen3.5-4B-Base NCRI 15.2; no filler; sealed 46.0
  185. 185 Gemma 2 9B NCRI 15.2; no filler; sealed 44.7
  186. 186 Qwen3.5-4B NCRI 15.2; no filler; sealed 44.3
  187. 187 Gemma 4 E2B NCRI 15.2; no filler; sealed 44.0
  188. 188 UI-TARS-1.5 NCRI 15.2; no filler; sealed 43.9
  189. 189 Mistral-Nemo-Base-2407 NCRI 15.2; no filler; sealed 43.2
  190. 190 Mistral NeMo NCRI 15.2; no filler; sealed 42.5
  191. 191 Llama 3.1-8B NCRI 15.2; no filler; sealed 40.6
  192. 192 Gemma 3n E4B NCRI 15.2; no filler; sealed 38.5
  193. 193 Gemma 3 4B NCRI 15.2; no filler; sealed 37.9
  194. 194 LLaMA-65B NCRI 15.2; no filler; sealed 37.5
  195. 195 Qwen3.5-2B NCRI 15.2; no filler; sealed 35.5
  196. 196 davinci-002 NCRI 15.2; no filler; sealed 32.6
  197. 197 GPT-5 nano NCRI 15.2; no filler; sealed 31.0
  198. 198 Qwen3.5-0.8B NCRI 15.2; no filler; sealed 24.9
  199. 199 Gemma 2 2B IT NCRI 15.2; no filler; sealed 24.1
  200. 200 Gemma 3 1B NCRI 15.2; no filler; sealed 16.1
  201. 201 LLaMA-7B NCRI 15.2; no filler; sealed 15.0
  202. 202 Gemma 3 270M NCRI 15.2; no filler; sealed -23.1

Sources: Official nocot-bench (NCRI) leaderboard and labs’ model cards

#EntryScore ± 95% CISettingDateSource
1 GPT-6 Astra 173.9 NCRI 15.2; dots@2048; placed Sep 2026 Official leaderboard
2 GPT-6.1 Sol 156.6 NCRI 15.2; dots@128; placed_after_release Sep 2026 Official leaderboard
3 Gemini 4 Argon 154.1 NCRI 15.2; qrep@512; placed_after_release Sep 2026 Official leaderboard
4 GPT-6.1 Terra 149.6 NCRI 15.2; qrep@512; placed_after_release Oct 2026 Official leaderboard
5 Gemini 3.8 Flash 129.7 NCRI 15.2; dots@128; placed Sep 2026 Official leaderboard
6 Claude Fable 5.1 129.6 NCRI 15.2; dots@128; placed Sep 2026 Official leaderboard
7 Gemini 3.7 Flash 127.6 NCRI 15.2; dots@32; sealed Aug 2026 Official leaderboard
8 Gemini 3.1 Pro 123.9 NCRI 15.2; count@512; sealed Feb 2026 Official leaderboard
9 GPT-6 Sol 122.2 NCRI 15.2; dots@128; placed_after_release Sep 2026 Official leaderboard
10 Claude Fable 5 121.5 NCRI 15.2; dots@32; sealed Jun 2026 Official leaderboard
11 GPT-5.6 Sol 119.5 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
12 Claude Opus 5.5 118.9 116.8 to 121.1 NCRI 15.2; dots@512; placed_after_release Sep 2026 Official leaderboard
13 Claude Opus 5 115.8 NCRI 15.2; dots@128; sealed Jul 2026 Official leaderboard
14 GPT-5.5 111.0 NCRI 15.2; qrep@512; sealed Apr 2026 Official leaderboard
15 Gemini 3.6 Flash 110.9 NCRI 15.2; count@512; sealed Jul 2026 Official leaderboard
16 Claude Opus 4.6 110.8 NCRI 15.2; qrep@512; sealed Feb 2026 Official leaderboard
17 Gemini 3 Flash 110.5 NCRI 15.2; count@512; sealed Dec 2025 Official leaderboard
18 Claude Opus 4.5 110.5 NCRI 15.2; dots@512; sealed Nov 2025 Official leaderboard
19 Claude Opus 4 109.0 NCRI 15.2; dots@512; sealed May 2025 Official leaderboard
20 Muse Spark 1.1 108.8 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
21 Claude Opus 4.1 107.5 NCRI 15.2; dots@128; sealed Aug 2025 Official leaderboard
22 Gemini 3.5 Flash 107.5 NCRI 15.2; dots@512; sealed May 2026 Official leaderboard
23 Kimi K3 105.6 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
24 Claude Opus 4.7 103.9 NCRI 15.2; dots@128; sealed Apr 2026 Official leaderboard
25 google/gemini-2.5-pro (snapshot unspecified) 102.9 NCRI 15.2; dots@512; sealed Official leaderboard
26 Kimi K3.1 Pro 102.8 100.8 to 105 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
27 Muse Spark 1.2 102.3 NCRI 15.2; dots@128; sealed Aug 2026 Official leaderboard
28 Claude Opus 4.8 102.2 NCRI 15.2; dots@128; sealed May 2026 Official leaderboard
29 Muse Spark 1.3 100.8 NCRI 15.2; dots@2048; placed Sep 2026 Official leaderboard
30 Claude Sonnet 4.5 100.8 NCRI 15.2; qrep@512; sealed Sep 2025 Official leaderboard
31 Muse Wonder 97.7 95.8 to 99.8 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
32 Claude Sonnet 4 96.5 NCRI 15.2; count@512; sealed May 2025 Official leaderboard
33 Muse Spark Open 95.7 ±2.1 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
34 GPT-5.6 Terra 95.1 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
35 Claude Sonnet 5.5 95.1 92.6 to 97.7 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
36 GPT-5.4 94.1 NCRI 15.2; dots@128; sealed Mar 2026 Official leaderboard
37 Tencent Hy4 94.0 ±2.1 NCRI 15.2; no filler; placed_after_release Nov 2026 Official leaderboard
38 Claude Sonnet 5 91.1 NCRI 15.2; dots@128; sealed Jun 2026 Official leaderboard
39 Claude Sonnet 4.6 90.0 NCRI 15.2; dots@512; sealed Feb 2026 Official leaderboard
40 Qwen 3.6 Max (Preview) 87.0 84.5 to 89 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
41 Grok 4.6 86.8 84.1 to 88.9 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
42 Gemini 3.1 Flash-Lite 86.5 83.8 to 88.7 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
43 Qwen3.8-2.4T-A95B 86.0 83.3 to 88.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
44 GPT-5.2 85.8 83.1 to 88 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
45 Qwen3.7-Max 85.8 83.2 to 88.1 NCRI 15.2; no filler; sealed May 2026 Official leaderboard
46 Qwen 3.8 Max 85.4 82.6 to 87.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
47 GPT-5.3 Codex 85.2 82.5 to 87.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
48 DeepSeek V4 Pro 0813 84.1 81.2 to 86.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
49 DeepSeek-V4-Pro 84.1 81.3 to 86.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
50 Qwen3-Max 83.3 80.7 to 85.5 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
51 Qwen3.7-Plus 83.1 80.2 to 85.2 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
52 Tencent Hy4 preview 82.8 79.9 to 85 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
53 LongCat-2.0 82.5 79.9 to 84.6 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
54 Qwen 3.6 Plus 82.2 79.3 to 84.3 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
55 Gemma 4 31B IT 81.7 78.6 to 83.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
56 DeepSeek V4.1 Flash 1017 81.6 ±2.2 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
57 DeepSeek V4.1 Flash 81.3 79 to 83.5 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
58 Grok 4.5 81.0 78.1 to 83.3 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
59 kwaipilot/kat-coder-pro-v2 80.8 78 to 82.8 NCRI 15.2; no filler; sealed Official leaderboard
60 Gemini 3.5 Flash-Lite 80.6 77.9 to 82.7 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
61 Qwen3.5-Plus 80.6 77.8 to 82.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
62 Kimi K2.5 80.6 77.7 to 82.8 NCRI 15.2; no filler; sealed Jan 2026 Official leaderboard
63 Qwen3.5 397B-A17B 80.0 77.1 to 81.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
64 Kimi K2.6 79.2 76.3 to 81.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
65 DeepSeek-V4-Flash-Base 79.1 76.1 to 81.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
66 google/gemini-2.5-flash (snapshot unspecified) 78.8 76 to 80.7 NCRI 15.2; no filler; sealed Official leaderboard
67 nvidia/nemotron-3-ultra-550b-a55b 78.3 75.4 to 80.3 NCRI 15.2; no filler; sealed Official leaderboard
68 Inkling 78.3 75.4 to 80.2 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
69 DeepSeek-V4-Flash 78.1 75.2 to 80.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
70 moonshotai/kimi-k2 (snapshot unspecified) 78.0 75 to 80.1 NCRI 15.2; no filler; sealed Official leaderboard
71 DeepSeek-V3.1-Terminus 77.6 74.7 to 79.9 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
72 GLM-5.2 77.3 74 to 79.6 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
73 GPT-6 Luna 77.3 74.7 to 79.1 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
74 Qwen3-VL-235B-A22B-Instruct 77.3 74.3 to 79.3 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
75 MiMo-V2.5-Pro 77.1 74.2 to 79.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
76 DeepSeek-V3.2-Exp 76.7 73.8 to 79 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
77 GPT-5.6 Luna 76.6 73.5 to 78.7 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
78 deepseek/deepseek-chat-v3.1 76.5 73.5 to 78.8 NCRI 15.2; no filler; sealed Official leaderboard
79 GLM-5.3-Flash 76.4 73.2 to 78.6 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
80 GLM-5 76.2 73.2 to 78.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
81 DeepSeek V4 Flash 0731 75.5 72.6 to 77.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
82 GLM-5.1 75.2 71.8 to 77.7 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
83 GPT-4.1 75.1 72 to 77.4 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
84 DeepSeek-V3.2 75.0 72 to 77.2 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
85 Qwen3-235B-A22B-Instruct Jul 2025 74.6 71.4 to 76.7 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
86 GLM-5.3 74.5 71.5 to 76.8 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
87 Qwen3.8 Flash 74.4 71.3 to 76.8 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
88 deepseek/deepseek-chat-v3-0324 74.4 71.5 to 76.5 NCRI 15.2; no filler; sealed Official leaderboard
89 openai/gpt-4o (snapshot unspecified) 74.2 71.3 to 76.3 NCRI 15.2; no filler; sealed Official leaderboard
90 GPT-4o Aug 2024 74.2 71.2 to 76.4 NCRI 15.2; no filler; sealed Aug 2024 Official leaderboard
91 Claude Haiku 4.5 74.2 71.1 to 76.4 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
92 Ling-2.6-1T 73.9 71 to 76.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
93 openai/gpt-4 (snapshot unspecified) 73.7 70.6 to 75.8 NCRI 15.2; no filler; sealed Official leaderboard
94 WizardLM-2 8x22B 73.2 70.2 to 75.3 NCRI 15.2; no filler; sealed Apr 2024 Official leaderboard
95 Qwen3-Coder-480B-A35B 73.1 69.7 to 75.5 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
96 deepseek/deepseek-chat 72.9 70.1 to 75.1 NCRI 15.2; no filler; sealed Official leaderboard
97 GPT-4 Jun 2023 72.8 69.7 to 75 NCRI 15.2; no filler; sealed Jun 2023 Official leaderboard
98 MiniMax-M3 72.6 69.2 to 74.7 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
99 Tencent Hy3 72.5 69.3 to 74.7 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
100 Mercury 3 72.3 69.9 to 74.5 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
101 Tencent Hy3 preview 72.3 69 to 74.5 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
102 GPT-4o May 2024 72.2 69.1 to 74.3 NCRI 15.2; no filler; sealed May 2024 Official leaderboard
103 GLM-4.7 71.3 68 to 73.5 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
104 openai/gpt-4-turbo (snapshot unspecified) 71.3 68.4 to 73.7 NCRI 15.2; no filler; sealed Official leaderboard
105 GLM-4.5 71.3 68.1 to 73.3 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
106 Qwen3-Coder-Plus 71.2 68 to 73.3 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
107 GLM-4.6 70.8 67.6 to 73.2 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
108 GPT-5.1 70.6 67.4 to 72.9 NCRI 15.2; no filler; sealed Nov 2025 Official leaderboard
109 Qwen3-235B-A22B 70.5 67.3 to 72.9 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
110 Llama 4 Maverick 70.5 67.5 to 72.7 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
111 GPT-5.4 Mini 70.5 67.3 to 72.4 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
112 GPT-4o Nov 2024 70.3 67.2 to 72.6 NCRI 15.2; no filler; sealed Nov 2024 Official leaderboard
113 Inkling-Small 70.2 66.7 to 72.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
114 DiffusionGemma 26B A4B 70.2 66.7 to 72.8 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
115 MiMo-V2.5 69.9 66.9 to 72.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
116 Gemma 4 26B A4B 69.5 66 to 71.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
117 Qwen3.5-122B-A10B 69.4 66.4 to 71.6 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
118 Grok 4.3 69.2 66 to 71.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
119 Qwen3-235B-A22B-Thinking Jul 2025 68.9 65.8 to 71 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
120 Qwen3-Next-80B-A3B-Instruct 68.4 65.2 to 70.5 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
121 MiniMax-M3.1-Flash 68.4 65.6 to 71.3 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
122 Claude Haiku 5.5 68.3 65.6 to 70.9 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
123 Muse Glimmer 67.9 64.3 to 70.2 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
124 Mistral Medium 3.5 67.5 64.1 to 69.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
125 Qwen3.6 27B 67.4 63.9 to 69.6 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
126 Grok 4.20 67.1 64 to 69.2 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
127 GPT-5 67.0 63.6 to 69.1 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
128 baidu/ernie-4.5-vl-424b-a47b 66.9 63.7 to 69 NCRI 15.2; no filler; sealed Official leaderboard
129 GLM-4.5V 66.7 63.3 to 69 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
130 Qwen3.5-27B 66.6 63 to 68.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
131 Ling-3.0-flash 66.3 63 to 68.7 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
132 nvidia/nemotron-3-super-120b-a12b 66.2 62.9 to 68.4 NCRI 15.2; no filler; sealed Official leaderboard
133 Qwen 3.8 27B 66.1 62.7 to 68.6 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
134 Mercury 2.5 65.8 62.4 to 68.1 NCRI 15.2; no filler; placed Sep 2026 Official leaderboard
135 inclusionai/ling-2.6-flash 65.5 62.4 to 67.9 NCRI 15.2; no filler; sealed Official leaderboard
136 Qwen3-Coder-Next 65.5 61.8 to 67.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
137 Solar Pro 4 65.3 62 to 67.7 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
138 Qwen2.5-VL-72B 65.2 61.7 to 67.6 NCRI 15.2; no filler; sealed Jan 2025 Official leaderboard
139 GPT-4.1 mini 65.0 61.4 to 67.4 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
140 MiniMax-M2.5 65.0 61.6 to 67.1 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
141 reflection/beam-1.0-501b-a23b 64.8 62.1 to 67.6 NCRI 15.2; no filler; placed_after_release Official leaderboard
142 nousresearch/hermes-3-llama-3.1-405b 64.6 61.2 to 66.9 NCRI 15.2; no filler; sealed Official leaderboard
143 Qwen3.5-35B-A3B-Base 64.5 61.1 to 66.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
144 google/gemini-2.5-flash-lite (snapshot unspecified) 64.5 61 to 66.8 NCRI 15.2; no filler; sealed Official leaderboard
145 MiniMax-M2.7 64.0 60.6 to 66.3 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
146 Qwen3-VL-32B-Instruct 64.0 60.5 to 66.7 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
147 Qwen3-Next-80B-A3B 63.8 60.4 to 66.1 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
148 sao10k/l3.3-euryale-70b 63.8 60.2 to 66.3 NCRI 15.2; no filler; sealed Official leaderboard
149 mistralai/mistral-small-2603 63.5 60.1 to 65.7 NCRI 15.2; no filler; sealed Official leaderboard
150 Qwen2.5-72B 63.5 60 to 66 NCRI 15.2; no filler; sealed Sep 2024 Official leaderboard
151 GLM-4.6V 63.4 59.7 to 65.9 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
152 Qwen 3.6 Flash 63.3 59.6 to 65.7 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
153 Qwen3.7 Flash 63.3 59.7 to 65.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
154 Hermes 4 405B 62.8 59.2 to 65.2 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
155 meta-llama/llama-3.1-70b-instruct 62.8 59.3 to 65.2 NCRI 15.2; no filler; sealed Official leaderboard
156 local/qwen2.5-32b-instruct 62.7 59.1 to 65.2 NCRI 15.2; no filler; sealed Official leaderboard
157 Llama 3.3 70B Instruct 62.6 59 to 64.8 NCRI 15.2; no filler; sealed Dec 2024 Official leaderboard
158 Laguna S 2.1 62.5 59.1 to 64.8 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
159 GLM-4.5-Air 62.4 58.8 to 64.8 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
160 Qwen3-30B-A3B-Instruct Jul 2025 62.1 58.5 to 64.6 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
161 Qwen 3.6 35B-A3B 61.6 58.3 to 64.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
162 Qwen3.5-Flash 61.4 57.8 to 63.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
163 amazon/nova-pro-v1 61.3 57.5 to 63.7 NCRI 15.2; no filler; sealed Official leaderboard
164 Qwen2.5-32B 61.1 57.6 to 63.5 NCRI 15.2; no filler; sealed Sep 2024 Official leaderboard
165 sao10k/l3.1-euryale-70b 61.0 57.2 to 63.4 NCRI 15.2; no filler; sealed Official leaderboard
166 Qwen2-72B 61.0 57.5 to 63.4 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
167 Qwen3.5-35B-A3B 60.1 56.3 to 62.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
168 Qwen2.5-Coder-32B-Instruct 60.0 56.1 to 62.5 NCRI 15.2; no filler; sealed Nov 2024 Official leaderboard
169 Qwen3-VL-30B-A3B-Instruct 59.9 56.1 to 62.6 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
170 Gemma 4 12B 59.7 55.6 to 62.4 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
171 Qwen3-Coder-30B-A3B-Instruct 59.6 55.9 to 62.3 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
172 Qwen3-Coder-Flash 59.5 55.7 to 61.9 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
173 Ripple-65B Preview 59.3 55 to 62.4 NCRI 15.2; no filler; placed_after_release Nov 2026 Official leaderboard
174 Qwen3-32B 59.2 55.2 to 61.8 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
175 thedrummer/cydonia-24b-v4.1 59.1 55.3 to 61.6 NCRI 15.2; no filler; sealed Official leaderboard
176 gpt-oss-120b 59.0 54.9 to 61.4 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
177 mistralai/ministral-14b-2512 58.7 55 to 61.2 NCRI 15.2; no filler; sealed Official leaderboard
178 Phi-4 58.3 54.6 to 61 NCRI 15.2; no filler; sealed Dec 2024 Official leaderboard
179 nousresearch/hermes-4-70b 57.9 54 to 60.4 NCRI 15.2; no filler; sealed Official leaderboard
180 Llama 4 Scout 57.7 53.9 to 60.2 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
181 Llama 3.1-70B 57.6 54 to 60.1 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
182 Laguna XS 2.1 57.5 53.3 to 59.9 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
183 Gemma 3 27B PT 57.5 53.7 to 59.9 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
184 Claude 3 Haiku 57.3 53.2 to 59.8 NCRI 15.2; no filler; sealed Mar 2024 Official leaderboard
185 local/mistral-small-24b-base-2501 56.2 52.2 to 58.6 NCRI 15.2; no filler; sealed Official leaderboard
186 Nemotron 3.5 Lightning 30B-A3B 56.1 52.2 to 58.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
187 Qwen3-VL-8B-Instruct 56.0 52.2 to 58.5 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
188 Gemma 3 27B 56.0 52 to 58.4 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
189 glm-4.7-flash 55.9 51.7 to 58.4 NCRI 15.2; no filler; sealed Jan 2026 Official leaderboard
190 GPT-4o mini 55.7 51.6 to 58.2 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
191 mistralai/mistral-small-24b-instruct-2501 55.4 51.6 to 57.8 NCRI 15.2; no filler; sealed Official leaderboard
192 openai/gpt-3.5-turbo (snapshot unspecified) 55.2 50.8 to 57.8 NCRI 15.2; no filler; sealed Official leaderboard
193 Mercury 2 55.1 51.1 to 57.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
194 Qwen3-14B 55.0 51.2 to 57.7 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
195 gpt-oss-20b 54.8 50.3 to 57.8 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
196 Olmo 3.1 32B Instruct 54.3 50.3 to 57 NCRI 15.2; no filler; placed Dec 2025 Official leaderboard
197 Qwen3.5-9B-Base 54.2 50 to 56.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
198 local/olmo-3-32b-dpo 54.0 50.1 to 56.8 NCRI 15.2; no filler; sealed Official leaderboard
199 mistralai/mistral-small-3.2-24b-instruct 53.8 49.4 to 56.5 NCRI 15.2; no filler; sealed Official leaderboard
200 local/olmo-3-32b-rl 53.8 49.9 to 56.4 NCRI 15.2; no filler; sealed Official leaderboard
201 local/olmo-3-32b-mid 53.3 49.5 to 55.7 NCRI 15.2; no filler; sealed Official leaderboard
202 Gemma 2 27B 53.3 49.2 to 55.7 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
203 local/olmo-3-32b-sft 52.8 48.6 to 55.4 NCRI 15.2; no filler; sealed Official leaderboard
204 google/gemma-2-27b-it 52.6 48.5 to 55.5 NCRI 15.2; no filler; sealed Official leaderboard
205 local/nanbeige42-r2 52.6 48.1 to 56.3 NCRI 15.2; no filler; placed_after_release Official leaderboard
206 local/qwen1.5-110b-chat 52.6 48.3 to 55.3 NCRI 15.2; no filler; sealed Official leaderboard
207 Qwen3.5-9B 52.4 47.8 to 55.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
208 local/ouro26-r4 51.7 46.1 to 56.6 NCRI 15.2; no filler; placed_after_release Official leaderboard
209 Qwen3-8B 51.4 46.8 to 54.3 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
210 mistralai/ministral-8b-2512 51.0 46.6 to 54 NCRI 15.2; no filler; sealed Official leaderboard
211 GPT-5 mini 50.7 45.7 to 53.6 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
212 Gemma 4 E4B 50.6 45.8 to 53.4 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
213 Qwen3-30B-A3B 50.4 45.3 to 53.5 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
214 GPT-4.1 nano 50.2 45.4 to 52.9 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
215 Gemma 3 12B PT 49.9 45.4 to 52.4 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
216 Qwen1.5-72B 49.7 45.4 to 52.7 NCRI 15.2; no filler; sealed Feb 2024 Official leaderboard
217 GPT-3.5 Turbo Instruct 49.6 45.2 to 51.9 NCRI 15.2; no filler; sealed Sep 2023 Official leaderboard
218 GPT-5.4 Nano 49.4 45.1 to 51.9 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
219 Gemma 3 12B 49.3 44.9 to 52.2 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
220 Hunyuan-A13B-Instruct 48.7 44.2 to 51.8 NCRI 15.2; no filler; sealed Jun 2025 Official leaderboard
221 qwen/qwen-2.5-7b-instruct 48.4 43.7 to 51.3 NCRI 15.2; no filler; sealed Official leaderboard
222 local/gemma-2-9b-it 47.5 42.9 to 50.6 NCRI 15.2; no filler; sealed Official leaderboard
223 Nemotron 3 Nano 30B-A3B 47.3 42 to 50.2 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
224 local/ouro14-r4 47.3 41.4 to 52.4 NCRI 15.2; no filler; placed_after_release Official leaderboard
225 mistralai/ministral-3b-2512 46.2 41.2 to 49.3 NCRI 15.2; no filler; sealed Official leaderboard
226 Qwen3.5-4B-Base 46.0 40.9 to 49.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
227 local/olmo-3-32b-pretrain 45.0 39.9 to 48 NCRI 15.2; no filler; sealed Official leaderboard
228 Gemma 2 9B 44.7 39.6 to 47.9 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
229 Qwen3.5-4B 44.3 39.2 to 48 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
230 Gemma 4 E2B 44.0 39.1 to 46.5 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
231 UI-TARS-1.5 43.9 38.9 to 47 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
232 Mistral-Nemo-Base-2407 43.2 38.4 to 46.1 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
233 thedrummer/unslopnemo-12b 42.8 37.7 to 45.4 NCRI 15.2; no filler; sealed Official leaderboard
234 Mistral NeMo 42.5 37.4 to 45.4 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
235 local/llama-3.1-8b-instruct-fp8 41.8 35.8 to 45.4 NCRI 15.2; no filler; sealed Official leaderboard
236 meta-llama/llama-3.1-8b-instruct 41.6 35.5 to 45 NCRI 15.2; no filler; sealed Official leaderboard
237 local/llama-3.1-8b-instruct 40.9 34.8 to 44.3 NCRI 15.2; no filler; sealed Official leaderboard
238 local/olmo3-32b-step260000 40.8 35.9 to 43.9 NCRI 15.2; no filler; placed Official leaderboard
239 local/llama-2-70b-hf 40.7 35.3 to 43.6 NCRI 15.2; no filler; sealed Official leaderboard
240 sao10k/l3-lunaris-8b 40.7 35.2 to 44.1 NCRI 15.2; no filler; sealed Official leaderboard
241 Llama 3.1-8B 40.6 35.1 to 43.8 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
242 local/llama-3-8b-instruct 39.2 33.9 to 42.4 NCRI 15.2; no filler; sealed Official leaderboard
243 Gemma 3n E4B 38.5 32.2 to 42.1 NCRI 15.2; no filler; sealed Jun 2025 Official leaderboard
244 local/olmo3-32b-step360000 38.5 33 to 41.8 NCRI 15.2; no filler; placed Official leaderboard
245 Gemma 3 4B 37.9 32 to 41.3 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
246 LLaMA-65B 37.5 31.7 to 39.9 NCRI 15.2; no filler; sealed Feb 2023 Official leaderboard
247 aion-labs/aion-rp-llama-3.1-8b 37.2 31.7 to 40.3 NCRI 15.2; no filler; sealed Official leaderboard
248 ibm-granite/granite-4.0-h-micro 37.2 31.1 to 40.8 NCRI 15.2; no filler; sealed Official leaderboard
249 local/olmo3-32b-step130000 36.2 30.5 to 39.5 NCRI 15.2; no filler; placed Official leaderboard
250 Qwen3.5-2B 35.5 29.1 to 38.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
251 local/olmo3-32b-step92000 34.5 28.4 to 37.4 NCRI 15.2; no filler; placed Official leaderboard
252 local/olmo3-32b-step182000 34.4 28.3 to 37.7 NCRI 15.2; no filler; sealed Official leaderboard
253 davinci-002 32.6 26 to 35.9 NCRI 15.2; no filler; sealed Mar 2022 Official leaderboard
254 local/olmo3-7b-step1413814 32.0 25.5 to 35.1 NCRI 15.2; no filler; sealed Official leaderboard
255 local/olmo3-32b-step65000 31.4 24.1 to 34.8 NCRI 15.2; no filler; placed Official leaderboard
256 GPT-5 nano 31.0 22.2 to 35.3 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
257 local/olmo3-32b-step46000 29.5 22.7 to 33.1 NCRI 15.2; no filler; sealed Official leaderboard
258 local/olmo3-32b-step33000 27.4 19.2 to 31.3 NCRI 15.2; no filler; placed Official leaderboard
259 local/gemma-7b-it 25.8 17.7 to 29.9 NCRI 15.2; no filler; sealed Official leaderboard
260 undi95/remm-slerp-l2-13b 25.7 16.3 to 30.1 NCRI 15.2; no filler; sealed Official leaderboard
261 Qwen3.5-0.8B 24.9 15.4 to 27.8 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
262 local/olmo3-7b-step360000 24.6 15.9 to 29.1 NCRI 15.2; no filler; sealed Official leaderboard
263 local/gemma-2-2b 24.3 15.7 to 28.3 NCRI 15.2; no filler; sealed Official leaderboard
264 Gemma 2 2B IT 24.1 14.5 to 28.7 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
265 local/talkie-1930-13b-it 23.9 18.8 to 28.3 NCRI 15.2; no filler; sealed Official leaderboard
266 gryphe/mythomax-l2-13b 23.8 13.2 to 29.4 NCRI 15.2; no filler; sealed Official leaderboard
267 local/olmo3-7b-step714000 23.2 12.7 to 27.7 NCRI 15.2; no filler; sealed Official leaderboard
268 local/olmo3-7b-step182000 22.7 13.1 to 27.4 NCRI 15.2; no filler; sealed Official leaderboard
269 local/olmo3-32b-step16000 21.5 12.5 to 26.3 NCRI 15.2; no filler; placed Official leaderboard
270 local/olmo3-32b-step23000 20.4 9.4 to 25.1 NCRI 15.2; no filler; placed Official leaderboard
271 local/olmo3-7b-step92000 18.9 7.7 to 24.1 NCRI 15.2; no filler; sealed Official leaderboard
272 local/llama-2-7b-hf 18.7 7.1 to 24.2 NCRI 15.2; no filler; sealed Official leaderboard
273 local/huginn-r12 16.8 2.9 to 25.9 NCRI 15.2; no filler; placed_after_release Official leaderboard
274 Gemma 3 1B 16.1 2 to 22.1 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
275 local/olmo3-32b-step12000 15.6 2.4 to 21 NCRI 15.2; no filler; placed Official leaderboard
276 LLaMA-7B 15.0 0.7 to 21.6 NCRI 15.2; no filler; sealed Feb 2023 Official leaderboard
277 local/gemma-2b-it 11.9 -5.3 to 18.5 NCRI 15.2; no filler; sealed Official leaderboard
278 local/talkie-web-13b-base 10.6 2.9 to 15.8 NCRI 15.2; no filler; sealed Official leaderboard
279 openai/babbage-002 6.6 -12.5 to 15.3 NCRI 15.2; no filler; sealed Official leaderboard
280 local/olmo3-7b-step46000 4.1 -14.4 to 13.3 NCRI 15.2; no filler; sealed Official leaderboard
281 local/olmo3-7b-step23000 3.1 -14.3 to 13 NCRI 15.2; no filler; sealed Official leaderboard
282 local/olmo3-7b-step12000 -8.3 -22.7 to 3.8 NCRI 15.2; no filler; sealed Official leaderboard
283 local/pythia-12b -8.6 ±13.5 NCRI 15.2; no filler; sealed Official leaderboard
284 local/olmo3-32b-step6000 -9.9 -23.2 to 4.2 NCRI 15.2; no filler; placed Official leaderboard
285 local/olmo3-7b-step6000 -11.5 ±12.8 NCRI 15.2; no filler; sealed Official leaderboard
286 local/talkie-1930-13b-base -12.4 ±12.1 NCRI 15.2; no filler; sealed Official leaderboard
287 Gemma 3 270M -23.1 -28.1 to -20.6 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
#ModelScore ± 95% CISettingDateSource
1 GPT-6 Astra 173.9 NCRI 15.2; dots@2048; placed Sep 2026 Official leaderboard
2 GPT-6.1 Sol 156.6 NCRI 15.2; dots@128; placed_after_release Sep 2026 Official leaderboard
3 Gemini 4 Argon 154.1 NCRI 15.2; qrep@512; placed_after_release Sep 2026 Official leaderboard
4 GPT-6.1 Terra 149.6 NCRI 15.2; qrep@512; placed_after_release Oct 2026 Official leaderboard
5 Gemini 3.8 Flash 129.7 NCRI 15.2; dots@128; placed Sep 2026 Official leaderboard
6 Claude Fable 5.1 129.6 NCRI 15.2; dots@128; placed Sep 2026 Official leaderboard
7 Gemini 3.7 Flash 127.6 NCRI 15.2; dots@32; sealed Aug 2026 Official leaderboard
8 Gemini 3.1 Pro 123.9 NCRI 15.2; count@512; sealed Feb 2026 Official leaderboard
9 GPT-6 Sol 122.2 NCRI 15.2; dots@128; placed_after_release Sep 2026 Official leaderboard
10 Claude Fable 5 121.5 NCRI 15.2; dots@32; sealed Jun 2026 Official leaderboard
11 GPT-5.6 Sol 119.5 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
12 Claude Opus 5.5 118.9 116.8 to 121.1 NCRI 15.2; dots@512; placed_after_release Sep 2026 Official leaderboard
13 Claude Opus 5 115.8 NCRI 15.2; dots@128; sealed Jul 2026 Official leaderboard
14 GPT-5.5 111.0 NCRI 15.2; qrep@512; sealed Apr 2026 Official leaderboard
15 Gemini 3.6 Flash 110.9 NCRI 15.2; count@512; sealed Jul 2026 Official leaderboard
16 Claude Opus 4.6 110.8 NCRI 15.2; qrep@512; sealed Feb 2026 Official leaderboard
17 Gemini 3 Flash 110.5 NCRI 15.2; count@512; sealed Dec 2025 Official leaderboard
18 Claude Opus 4.5 110.5 NCRI 15.2; dots@512; sealed Nov 2025 Official leaderboard
19 Claude Opus 4 109.0 NCRI 15.2; dots@512; sealed May 2025 Official leaderboard
20 Muse Spark 1.1 108.8 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
21 Claude Opus 4.1 107.5 NCRI 15.2; dots@128; sealed Aug 2025 Official leaderboard
22 Gemini 3.5 Flash 107.5 NCRI 15.2; dots@512; sealed May 2026 Official leaderboard
23 Kimi K3 105.6 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
24 Claude Opus 4.7 103.9 NCRI 15.2; dots@128; sealed Apr 2026 Official leaderboard
25 Kimi K3.1 Pro 102.8 100.8 to 105 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
26 Muse Spark 1.2 102.3 NCRI 15.2; dots@128; sealed Aug 2026 Official leaderboard
27 Claude Opus 4.8 102.2 NCRI 15.2; dots@128; sealed May 2026 Official leaderboard
28 Muse Spark 1.3 100.8 NCRI 15.2; dots@2048; placed Sep 2026 Official leaderboard
29 Claude Sonnet 4.5 100.8 NCRI 15.2; qrep@512; sealed Sep 2025 Official leaderboard
30 Muse Wonder 97.7 95.8 to 99.8 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
31 Claude Sonnet 4 96.5 NCRI 15.2; count@512; sealed May 2025 Official leaderboard
32 Muse Spark Open 95.7 ±2.1 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
33 GPT-5.6 Terra 95.1 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
34 Claude Sonnet 5.5 95.1 92.6 to 97.7 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
35 GPT-5.4 94.1 NCRI 15.2; dots@128; sealed Mar 2026 Official leaderboard
36 Tencent Hy4 94.0 ±2.1 NCRI 15.2; no filler; placed_after_release Nov 2026 Official leaderboard
37 Claude Sonnet 5 91.1 NCRI 15.2; dots@128; sealed Jun 2026 Official leaderboard
38 Claude Sonnet 4.6 90.0 NCRI 15.2; dots@512; sealed Feb 2026 Official leaderboard
39 Qwen 3.6 Max (Preview) 87.0 84.5 to 89 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
40 Grok 4.6 86.8 84.1 to 88.9 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
41 Gemini 3.1 Flash-Lite 86.5 83.8 to 88.7 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
42 Qwen3.8-2.4T-A95B 86.0 83.3 to 88.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
43 GPT-5.2 85.8 83.1 to 88 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
44 Qwen3.7-Max 85.8 83.2 to 88.1 NCRI 15.2; no filler; sealed May 2026 Official leaderboard
45 Qwen 3.8 Max 85.4 82.6 to 87.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
46 GPT-5.3 Codex 85.2 82.5 to 87.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
47 DeepSeek V4 Pro 0813 84.1 81.2 to 86.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
48 DeepSeek-V4-Pro 84.1 81.3 to 86.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
49 Qwen3-Max 83.3 80.7 to 85.5 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
50 Qwen3.7-Plus 83.1 80.2 to 85.2 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
51 Tencent Hy4 preview 82.8 79.9 to 85 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
52 LongCat-2.0 82.5 79.9 to 84.6 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
53 Qwen 3.6 Plus 82.2 79.3 to 84.3 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
54 Gemma 4 31B IT 81.7 78.6 to 83.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
55 DeepSeek V4.1 Flash 1017 81.6 ±2.2 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
56 DeepSeek V4.1 Flash 81.3 79 to 83.5 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
57 Grok 4.5 81.0 78.1 to 83.3 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
58 Gemini 3.5 Flash-Lite 80.6 77.9 to 82.7 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
59 Qwen3.5-Plus 80.6 77.8 to 82.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
60 Kimi K2.5 80.6 77.7 to 82.8 NCRI 15.2; no filler; sealed Jan 2026 Official leaderboard
61 Qwen3.5 397B-A17B 80.0 77.1 to 81.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
62 Kimi K2.6 79.2 76.3 to 81.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
63 DeepSeek-V4-Flash-Base 79.1 76.1 to 81.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
64 Inkling 78.3 75.4 to 80.2 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
65 DeepSeek-V4-Flash 78.1 75.2 to 80.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
66 DeepSeek-V3.1-Terminus 77.6 74.7 to 79.9 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
67 GLM-5.2 77.3 74 to 79.6 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
68 GPT-6 Luna 77.3 74.7 to 79.1 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
69 Qwen3-VL-235B-A22B-Instruct 77.3 74.3 to 79.3 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
70 MiMo-V2.5-Pro 77.1 74.2 to 79.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
71 DeepSeek-V3.2-Exp 76.7 73.8 to 79 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
72 GPT-5.6 Luna 76.6 73.5 to 78.7 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
73 GLM-5.3-Flash 76.4 73.2 to 78.6 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
74 GLM-5 76.2 73.2 to 78.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
75 DeepSeek V4 Flash 0731 75.5 72.6 to 77.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
76 GLM-5.1 75.2 71.8 to 77.7 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
77 GPT-4.1 75.1 72 to 77.4 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
78 DeepSeek-V3.2 75.0 72 to 77.2 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
79 Qwen3-235B-A22B-Instruct Jul 2025 74.6 71.4 to 76.7 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
80 GLM-5.3 74.5 71.5 to 76.8 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
81 Qwen3.8 Flash 74.4 71.3 to 76.8 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
82 GPT-4o Aug 2024 74.2 71.2 to 76.4 NCRI 15.2; no filler; sealed Aug 2024 Official leaderboard
83 Claude Haiku 4.5 74.2 71.1 to 76.4 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
84 Ling-2.6-1T 73.9 71 to 76.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
85 WizardLM-2 8x22B 73.2 70.2 to 75.3 NCRI 15.2; no filler; sealed Apr 2024 Official leaderboard
86 Qwen3-Coder-480B-A35B 73.1 69.7 to 75.5 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
87 GPT-4 Jun 2023 72.8 69.7 to 75 NCRI 15.2; no filler; sealed Jun 2023 Official leaderboard
88 MiniMax-M3 72.6 69.2 to 74.7 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
89 Tencent Hy3 72.5 69.3 to 74.7 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
90 Mercury 3 72.3 69.9 to 74.5 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
91 Tencent Hy3 preview 72.3 69 to 74.5 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
92 GPT-4o May 2024 72.2 69.1 to 74.3 NCRI 15.2; no filler; sealed May 2024 Official leaderboard
93 GLM-4.7 71.3 68 to 73.5 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
94 GLM-4.5 71.3 68.1 to 73.3 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
95 Qwen3-Coder-Plus 71.2 68 to 73.3 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
96 GLM-4.6 70.8 67.6 to 73.2 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
97 GPT-5.1 70.6 67.4 to 72.9 NCRI 15.2; no filler; sealed Nov 2025 Official leaderboard
98 Qwen3-235B-A22B 70.5 67.3 to 72.9 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
99 Llama 4 Maverick 70.5 67.5 to 72.7 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
100 GPT-5.4 Mini 70.5 67.3 to 72.4 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
101 GPT-4o Nov 2024 70.3 67.2 to 72.6 NCRI 15.2; no filler; sealed Nov 2024 Official leaderboard
102 Inkling-Small 70.2 66.7 to 72.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
103 DiffusionGemma 26B A4B 70.2 66.7 to 72.8 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
104 MiMo-V2.5 69.9 66.9 to 72.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
105 Gemma 4 26B A4B 69.5 66 to 71.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
106 Qwen3.5-122B-A10B 69.4 66.4 to 71.6 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
107 Grok 4.3 69.2 66 to 71.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
108 Qwen3-235B-A22B-Thinking Jul 2025 68.9 65.8 to 71 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
109 Qwen3-Next-80B-A3B-Instruct 68.4 65.2 to 70.5 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
110 MiniMax-M3.1-Flash 68.4 65.6 to 71.3 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
111 Claude Haiku 5.5 68.3 65.6 to 70.9 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
112 Muse Glimmer 67.9 64.3 to 70.2 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
113 Mistral Medium 3.5 67.5 64.1 to 69.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
114 Qwen3.6 27B 67.4 63.9 to 69.6 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
115 Grok 4.20 67.1 64 to 69.2 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
116 GPT-5 67.0 63.6 to 69.1 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
117 GLM-4.5V 66.7 63.3 to 69 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
118 Qwen3.5-27B 66.6 63 to 68.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
119 Ling-3.0-flash 66.3 63 to 68.7 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
120 Qwen 3.8 27B 66.1 62.7 to 68.6 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
121 Mercury 2.5 65.8 62.4 to 68.1 NCRI 15.2; no filler; placed Sep 2026 Official leaderboard
122 Qwen3-Coder-Next 65.5 61.8 to 67.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
123 Solar Pro 4 65.3 62 to 67.7 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
124 Qwen2.5-VL-72B 65.2 61.7 to 67.6 NCRI 15.2; no filler; sealed Jan 2025 Official leaderboard
125 GPT-4.1 mini 65.0 61.4 to 67.4 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
126 MiniMax-M2.5 65.0 61.6 to 67.1 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
127 Qwen3.5-35B-A3B-Base 64.5 61.1 to 66.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
128 MiniMax-M2.7 64.0 60.6 to 66.3 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
129 Qwen3-VL-32B-Instruct 64.0 60.5 to 66.7 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
130 Qwen3-Next-80B-A3B 63.8 60.4 to 66.1 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
131 Qwen2.5-72B 63.5 60 to 66 NCRI 15.2; no filler; sealed Sep 2024 Official leaderboard
132 GLM-4.6V 63.4 59.7 to 65.9 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
133 Qwen 3.6 Flash 63.3 59.6 to 65.7 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
134 Qwen3.7 Flash 63.3 59.7 to 65.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
135 Hermes 4 405B 62.8 59.2 to 65.2 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
136 Llama 3.3 70B Instruct 62.6 59 to 64.8 NCRI 15.2; no filler; sealed Dec 2024 Official leaderboard
137 Laguna S 2.1 62.5 59.1 to 64.8 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
138 GLM-4.5-Air 62.4 58.8 to 64.8 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
139 Qwen3-30B-A3B-Instruct Jul 2025 62.1 58.5 to 64.6 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
140 Qwen 3.6 35B-A3B 61.6 58.3 to 64.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
141 Qwen3.5-Flash 61.4 57.8 to 63.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
142 Qwen2.5-32B 61.1 57.6 to 63.5 NCRI 15.2; no filler; sealed Sep 2024 Official leaderboard
143 Qwen2-72B 61.0 57.5 to 63.4 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
144 Qwen3.5-35B-A3B 60.1 56.3 to 62.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
145 Qwen2.5-Coder-32B-Instruct 60.0 56.1 to 62.5 NCRI 15.2; no filler; sealed Nov 2024 Official leaderboard
146 Qwen3-VL-30B-A3B-Instruct 59.9 56.1 to 62.6 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
147 Gemma 4 12B 59.7 55.6 to 62.4 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
148 Qwen3-Coder-30B-A3B-Instruct 59.6 55.9 to 62.3 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
149 Qwen3-Coder-Flash 59.5 55.7 to 61.9 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
150 Ripple-65B Preview 59.3 55 to 62.4 NCRI 15.2; no filler; placed_after_release Nov 2026 Official leaderboard
151 Qwen3-32B 59.2 55.2 to 61.8 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
152 gpt-oss-120b 59.0 54.9 to 61.4 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
153 Phi-4 58.3 54.6 to 61 NCRI 15.2; no filler; sealed Dec 2024 Official leaderboard
154 Llama 4 Scout 57.7 53.9 to 60.2 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
155 Llama 3.1-70B 57.6 54 to 60.1 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
156 Laguna XS 2.1 57.5 53.3 to 59.9 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
157 Gemma 3 27B PT 57.5 53.7 to 59.9 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
158 Claude 3 Haiku 57.3 53.2 to 59.8 NCRI 15.2; no filler; sealed Mar 2024 Official leaderboard
159 Nemotron 3.5 Lightning 30B-A3B 56.1 52.2 to 58.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
160 Qwen3-VL-8B-Instruct 56.0 52.2 to 58.5 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
161 Gemma 3 27B 56.0 52 to 58.4 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
162 glm-4.7-flash 55.9 51.7 to 58.4 NCRI 15.2; no filler; sealed Jan 2026 Official leaderboard
163 GPT-4o mini 55.7 51.6 to 58.2 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
164 Mercury 2 55.1 51.1 to 57.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
165 Qwen3-14B 55.0 51.2 to 57.7 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
166 gpt-oss-20b 54.8 50.3 to 57.8 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
167 Olmo 3.1 32B Instruct 54.3 50.3 to 57 NCRI 15.2; no filler; placed Dec 2025 Official leaderboard
168 Qwen3.5-9B-Base 54.2 50 to 56.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
169 Gemma 2 27B 53.3 49.2 to 55.7 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
170 Qwen3.5-9B 52.4 47.8 to 55.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
171 Qwen3-8B 51.4 46.8 to 54.3 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
172 GPT-5 mini 50.7 45.7 to 53.6 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
173 Gemma 4 E4B 50.6 45.8 to 53.4 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
174 Qwen3-30B-A3B 50.4 45.3 to 53.5 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
175 GPT-4.1 nano 50.2 45.4 to 52.9 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
176 Gemma 3 12B PT 49.9 45.4 to 52.4 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
177 Qwen1.5-72B 49.7 45.4 to 52.7 NCRI 15.2; no filler; sealed Feb 2024 Official leaderboard
178 GPT-3.5 Turbo Instruct 49.6 45.2 to 51.9 NCRI 15.2; no filler; sealed Sep 2023 Official leaderboard
179 GPT-5.4 Nano 49.4 45.1 to 51.9 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
180 Gemma 3 12B 49.3 44.9 to 52.2 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
181 Hunyuan-A13B-Instruct 48.7 44.2 to 51.8 NCRI 15.2; no filler; sealed Jun 2025 Official leaderboard
182 Nemotron 3 Nano 30B-A3B 47.3 42 to 50.2 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
183 Qwen3.5-4B-Base 46.0 40.9 to 49.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
184 Gemma 2 9B 44.7 39.6 to 47.9 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
185 Qwen3.5-4B 44.3 39.2 to 48 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
186 Gemma 4 E2B 44.0 39.1 to 46.5 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
187 UI-TARS-1.5 43.9 38.9 to 47 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
188 Mistral-Nemo-Base-2407 43.2 38.4 to 46.1 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
189 Mistral NeMo 42.5 37.4 to 45.4 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
190 Llama 3.1-8B 40.6 35.1 to 43.8 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
191 Gemma 3n E4B 38.5 32.2 to 42.1 NCRI 15.2; no filler; sealed Jun 2025 Official leaderboard
192 Gemma 3 4B 37.9 32 to 41.3 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
193 LLaMA-65B 37.5 31.7 to 39.9 NCRI 15.2; no filler; sealed Feb 2023 Official leaderboard
194 Qwen3.5-2B 35.5 29.1 to 38.7 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
195 davinci-002 32.6 26 to 35.9 NCRI 15.2; no filler; sealed Mar 2022 Official leaderboard
196 GPT-5 nano 31.0 22.2 to 35.3 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
197 Qwen3.5-0.8B 24.9 15.4 to 27.8 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
198 Gemma 2 2B IT 24.1 14.5 to 28.7 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
199 Gemma 3 1B 16.1 2 to 22.1 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
200 LLaMA-7B 15.0 0.7 to 21.6 NCRI 15.2; no filler; sealed Feb 2023 Official leaderboard
201 Gemma 3 270M -23.1 -28.1 to -20.6 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-6 Astra 173.9 NCRI 15.2; dots@2048; placed Sep 2026 Official leaderboard
2 GPT-6.2 Astra 171.0 Nov 2026 Lab model card
3 GPT-6.1 Sol 156.6 NCRI 15.2; dots@128; placed_after_release Sep 2026 Official leaderboard
4 Gemini 4 Argon 154.1 NCRI 15.2; qrep@512; placed_after_release Sep 2026 Official leaderboard
5 GPT-6.1 Terra 149.6 NCRI 15.2; qrep@512; placed_after_release Oct 2026 Official leaderboard
6 Gemini 3.8 Flash 129.7 NCRI 15.2; dots@128; placed Sep 2026 Official leaderboard
7 Claude Fable 5.1 129.6 NCRI 15.2; dots@128; placed Sep 2026 Official leaderboard
8 Gemini 3.7 Flash 127.6 NCRI 15.2; dots@32; sealed Aug 2026 Official leaderboard
9 Gemini 3.1 Pro 123.9 NCRI 15.2; count@512; sealed Feb 2026 Official leaderboard
10 GPT-6 Sol 122.2 NCRI 15.2; dots@128; placed_after_release Sep 2026 Official leaderboard
11 Claude Fable 5 121.5 NCRI 15.2; dots@32; sealed Jun 2026 Official leaderboard
12 GPT-5.6 Sol 119.5 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
13 Claude Opus 5.5 118.9 NCRI 15.2; dots@512; placed_after_release Sep 2026 Official leaderboard
14 Claude Opus 5 115.8 NCRI 15.2; dots@128; sealed Jul 2026 Official leaderboard
15 GPT-5.5 111.0 NCRI 15.2; qrep@512; sealed Apr 2026 Official leaderboard
16 Gemini 3.6 Flash 110.9 NCRI 15.2; count@512; sealed Jul 2026 Official leaderboard
17 Claude Opus 4.6 110.8 NCRI 15.2; qrep@512; sealed Feb 2026 Official leaderboard
18 Gemini 3 Flash 110.5 NCRI 15.2; count@512; sealed Dec 2025 Official leaderboard
19 Claude Opus 4.5 110.5 NCRI 15.2; dots@512; sealed Nov 2025 Official leaderboard
20 Claude Opus 4 109.0 NCRI 15.2; dots@512; sealed May 2025 Official leaderboard
21 Muse Spark 1.1 108.8 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
22 Claude Opus 4.1 107.5 NCRI 15.2; dots@128; sealed Aug 2025 Official leaderboard
23 Gemini 3.5 Flash 107.5 NCRI 15.2; dots@512; sealed May 2026 Official leaderboard
24 Kimi K3 105.6 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
25 Claude Opus 4.7 103.9 NCRI 15.2; dots@128; sealed Apr 2026 Official leaderboard
26 Kimi K3.1 Pro 102.8 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
27 Muse Spark 1.2 102.3 NCRI 15.2; dots@128; sealed Aug 2026 Official leaderboard
28 Claude Opus 4.8 102.2 NCRI 15.2; dots@128; sealed May 2026 Official leaderboard
29 Muse Spark 1.3 100.8 NCRI 15.2; dots@2048; placed Sep 2026 Official leaderboard
30 Claude Sonnet 4.5 100.8 NCRI 15.2; qrep@512; sealed Sep 2025 Official leaderboard
31 Muse Wonder 97.7 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
32 Claude Sonnet 4 96.5 NCRI 15.2; count@512; sealed May 2025 Official leaderboard
33 Muse Spark Open 95.7 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
34 GPT-5.6 Terra 95.1 NCRI 15.2; qrep@512; sealed Jul 2026 Official leaderboard
35 Claude Sonnet 5.5 95.1 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
36 GPT-5.4 94.1 NCRI 15.2; dots@128; sealed Mar 2026 Official leaderboard
37 Tencent Hy4 94.0 NCRI 15.2; no filler; placed_after_release Nov 2026 Official leaderboard
38 Claude Sonnet 5 91.1 NCRI 15.2; dots@128; sealed Jun 2026 Official leaderboard
39 Claude Sonnet 4.6 90.0 NCRI 15.2; dots@512; sealed Feb 2026 Official leaderboard
40 Qwen 3.6 Max (Preview) 87.0 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
41 Grok 4.6 86.8 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
42 Gemini 3.1 Flash-Lite 86.5 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
43 Qwen3.8-2.4T-A95B 86.0 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
44 GPT-5.2 85.8 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
45 Qwen3.7-Max 85.8 NCRI 15.2; no filler; sealed May 2026 Official leaderboard
46 Qwen 3.8 Max 85.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
47 GPT-5.3 Codex 85.2 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
48 DeepSeek V4 Pro 0813 84.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
49 DeepSeek-V4-Pro 84.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
50 Qwen3-Max 83.3 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
51 Qwen3.7-Plus 83.1 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
52 Tencent Hy4 preview 82.8 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
53 LongCat-2.0 82.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
54 Qwen 3.6 Plus 82.2 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
55 Gemma 4 31B IT 81.7 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
56 DeepSeek V4.1 Flash 1017 81.6 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
57 DeepSeek V4.1 Flash 81.3 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
58 Grok 4.5 81.0 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
59 Gemini 3.5 Flash-Lite 80.6 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
60 Qwen3.5-Plus 80.6 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
61 Kimi K2.5 80.6 NCRI 15.2; no filler; sealed Jan 2026 Official leaderboard
62 Qwen3.5 397B-A17B 80.0 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
63 Kimi K2.6 79.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
64 DeepSeek-V4-Flash-Base 79.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
65 Inkling 78.3 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
66 DeepSeek-V4-Flash 78.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
67 DeepSeek-V3.1-Terminus 77.6 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
68 GLM-5.2 77.3 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
69 GPT-6 Luna 77.3 NCRI 15.2; no filler; placed_after_release Sep 2026 Official leaderboard
70 Qwen3-VL-235B-A22B-Instruct 77.3 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
71 MiMo-V2.5-Pro 77.1 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
72 DeepSeek-V3.2-Exp 76.7 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
73 GPT-5.6 Luna 76.6 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
74 GLM-5.3-Flash 76.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
75 GLM-5 76.2 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
76 DeepSeek V4 Flash 0731 75.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
77 GLM-5.1 75.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
78 GPT-4.1 75.1 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
79 DeepSeek-V3.2 75.0 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
80 Qwen3-235B-A22B-Instruct Jul 2025 74.6 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
81 GLM-5.3 74.5 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
82 Qwen3.8 Flash 74.4 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
83 GPT-4o Aug 2024 74.2 NCRI 15.2; no filler; sealed Aug 2024 Official leaderboard
84 Claude Haiku 4.5 74.2 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
85 Ling-2.6-1T 73.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
86 WizardLM-2 8x22B 73.2 NCRI 15.2; no filler; sealed Apr 2024 Official leaderboard
87 Qwen3-Coder-480B-A35B 73.1 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
88 GPT-4 Jun 2023 72.8 NCRI 15.2; no filler; sealed Jun 2023 Official leaderboard
89 MiniMax-M3 72.6 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
90 Tencent Hy3 72.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
91 Mercury 3 72.3 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
92 Tencent Hy3 preview 72.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
93 GPT-4o May 2024 72.2 NCRI 15.2; no filler; sealed May 2024 Official leaderboard
94 GLM-4.7 71.3 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
95 GLM-4.5 71.3 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
96 Qwen3-Coder-Plus 71.2 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
97 GLM-4.6 70.8 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
98 GPT-5.1 70.6 NCRI 15.2; no filler; sealed Nov 2025 Official leaderboard
99 Qwen3-235B-A22B 70.5 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
100 Llama 4 Maverick 70.5 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
101 GPT-5.4 Mini 70.5 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
102 GPT-4o Nov 2024 70.3 NCRI 15.2; no filler; sealed Nov 2024 Official leaderboard
103 Inkling-Small 70.2 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
104 DiffusionGemma 26B A4B 70.2 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
105 MiMo-V2.5 69.9 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
106 Gemma 4 26B A4B 69.5 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
107 Qwen3.5-122B-A10B 69.4 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
108 Grok 4.3 69.2 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
109 Qwen3-235B-A22B-Thinking Jul 2025 68.9 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
110 Qwen3-Next-80B-A3B-Instruct 68.4 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
111 MiniMax-M3.1-Flash 68.4 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
112 Claude Haiku 5.5 68.3 NCRI 15.2; no filler; placed_after_release Oct 2026 Official leaderboard
113 Muse Glimmer 67.9 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
114 Mistral Medium 3.5 67.5 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
115 Qwen3.6 27B 67.4 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
116 Grok 4.20 67.1 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
117 GPT-5 67.0 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
118 GLM-4.5V 66.7 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
119 Qwen3.5-27B 66.6 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
120 Ling-3.0-flash 66.3 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
121 Qwen 3.8 27B 66.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
122 Mercury 2.5 65.8 NCRI 15.2; no filler; placed Sep 2026 Official leaderboard
123 Qwen3-Coder-Next 65.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
124 Solar Pro 4 65.3 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
125 Qwen2.5-VL-72B 65.2 NCRI 15.2; no filler; sealed Jan 2025 Official leaderboard
126 GPT-4.1 mini 65.0 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
127 MiniMax-M2.5 65.0 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
128 Qwen3.5-35B-A3B-Base 64.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
129 MiniMax-M2.7 64.0 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
130 Qwen3-VL-32B-Instruct 64.0 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
131 Qwen3-Next-80B-A3B 63.8 NCRI 15.2; no filler; sealed Sep 2025 Official leaderboard
132 Qwen2.5-72B 63.5 NCRI 15.2; no filler; sealed Sep 2024 Official leaderboard
133 GLM-4.6V 63.4 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
134 Qwen 3.6 Flash 63.3 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
135 Qwen3.7 Flash 63.3 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
136 Hermes 4 405B 62.8 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
137 Llama 3.3 70B Instruct 62.6 NCRI 15.2; no filler; sealed Dec 2024 Official leaderboard
138 Laguna S 2.1 62.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
139 GLM-4.5-Air 62.4 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
140 Qwen3-30B-A3B-Instruct Jul 2025 62.1 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
141 Qwen 3.6 35B-A3B 61.6 NCRI 15.2; no filler; sealed Apr 2026 Official leaderboard
142 Qwen3.5-Flash 61.4 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
143 Qwen2.5-32B 61.1 NCRI 15.2; no filler; sealed Sep 2024 Official leaderboard
144 Qwen2-72B 61.0 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
145 Qwen3.5-35B-A3B 60.1 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
146 Qwen2.5-Coder-32B-Instruct 60.0 NCRI 15.2; no filler; sealed Nov 2024 Official leaderboard
147 Qwen3-VL-30B-A3B-Instruct 59.9 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
148 Gemma 4 12B 59.7 NCRI 15.2; no filler; sealed Jun 2026 Official leaderboard
149 Qwen3-Coder-30B-A3B-Instruct 59.6 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
150 Qwen3-Coder-Flash 59.5 NCRI 15.2; no filler; sealed Jul 2025 Official leaderboard
151 Ripple-65B Preview 59.3 NCRI 15.2; no filler; placed_after_release Nov 2026 Official leaderboard
152 Qwen3-32B 59.2 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
153 gpt-oss-120b 59.0 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
154 Phi-4 58.3 NCRI 15.2; no filler; sealed Dec 2024 Official leaderboard
155 Llama 4 Scout 57.7 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
156 Llama 3.1-70B 57.6 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
157 Laguna XS 2.1 57.5 NCRI 15.2; no filler; sealed Jul 2026 Official leaderboard
158 Gemma 3 27B PT 57.5 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
159 Claude 3 Haiku 57.3 NCRI 15.2; no filler; sealed Mar 2024 Official leaderboard
160 Nemotron 3.5 Lightning 30B-A3B 56.1 NCRI 15.2; no filler; sealed Aug 2026 Official leaderboard
161 Qwen3-VL-8B-Instruct 56.0 NCRI 15.2; no filler; sealed Oct 2025 Official leaderboard
162 Gemma 3 27B 56.0 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
163 glm-4.7-flash 55.9 NCRI 15.2; no filler; sealed Jan 2026 Official leaderboard
164 GPT-4o mini 55.7 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
165 Mercury 2 55.1 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
166 Qwen3-14B 55.0 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
167 gpt-oss-20b 54.8 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
168 Olmo 3.1 32B Instruct 54.3 NCRI 15.2; no filler; placed Dec 2025 Official leaderboard
169 Qwen3.5-9B-Base 54.2 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
170 Gemma 2 27B 53.3 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
171 Qwen3.5-9B 52.4 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
172 Qwen3-8B 51.4 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
173 GPT-5 mini 50.7 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
174 Gemma 4 E4B 50.6 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
175 Qwen3-30B-A3B 50.4 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
176 GPT-4.1 nano 50.2 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
177 Gemma 3 12B PT 49.9 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
178 Qwen1.5-72B 49.7 NCRI 15.2; no filler; sealed Feb 2024 Official leaderboard
179 GPT-3.5 Turbo Instruct 49.6 NCRI 15.2; no filler; sealed Sep 2023 Official leaderboard
180 GPT-5.4 Nano 49.4 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
181 Gemma 3 12B 49.3 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
182 Hunyuan-A13B-Instruct 48.7 NCRI 15.2; no filler; sealed Jun 2025 Official leaderboard
183 Nemotron 3 Nano 30B-A3B 47.3 NCRI 15.2; no filler; sealed Dec 2025 Official leaderboard
184 Qwen3.5-4B-Base 46.0 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
185 Gemma 2 9B 44.7 NCRI 15.2; no filler; sealed Jun 2024 Official leaderboard
186 Qwen3.5-4B 44.3 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
187 Gemma 4 E2B 44.0 NCRI 15.2; no filler; sealed Mar 2026 Official leaderboard
188 UI-TARS-1.5 43.9 NCRI 15.2; no filler; sealed Apr 2025 Official leaderboard
189 Mistral-Nemo-Base-2407 43.2 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
190 Mistral NeMo 42.5 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
191 Llama 3.1-8B 40.6 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
192 Gemma 3n E4B 38.5 NCRI 15.2; no filler; sealed Jun 2025 Official leaderboard
193 Gemma 3 4B 37.9 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
194 LLaMA-65B 37.5 NCRI 15.2; no filler; sealed Feb 2023 Official leaderboard
195 Qwen3.5-2B 35.5 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
196 davinci-002 32.6 NCRI 15.2; no filler; sealed Mar 2022 Official leaderboard
197 GPT-5 nano 31.0 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard
198 Qwen3.5-0.8B 24.9 NCRI 15.2; no filler; sealed Feb 2026 Official leaderboard
199 Gemma 2 2B IT 24.1 NCRI 15.2; no filler; sealed Jul 2024 Official leaderboard
200 Gemma 3 1B 16.1 NCRI 15.2; no filler; sealed Mar 2025 Official leaderboard
201 LLaMA-7B 15.0 NCRI 15.2; no filler; sealed Feb 2023 Official leaderboard
202 Gemma 3 270M -23.1 NCRI 15.2; no filler; sealed Aug 2025 Official leaderboard

Labs report scores for 2 models in their cards. Switch the scores above to rank them together.

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-6 Astra
NCRI 15.2; … NCRI 15.2; …
173.9
GPT-6.1 Sol
NCRI 15.2; … NCRI 15.2; …
156.6
Gemini 4 Argon
NCRI 15.2; …
154.1
GPT-6.1 Terra
NCRI 15.2; … NCRI 15.2; …
149.6
Gemini 3.8 Flash
NCRI 15.2; …
129.7
Claude Fable 5.1
NCRI 15.2; …
129.6
Gemini 3.7 Flash
NCRI 15.2; …
127.6
Gemini 3.1 Pro
NCRI 15.2; … NCRI 15.2; …
123.9
GPT-6 Sol
NCRI 15.2; … NCRI 15.2; …
122.2
Claude Fable 5
NCRI 15.2; …
121.5
GPT-5.6 Sol
NCRI 15.2; … NCRI 15.2; …
119.5
Claude Opus 5
NCRI 15.2; …
115.8

Source: Official nocot-bench (NCRI) leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6 Astra NCRI 15.2; dots@2048; placed 173.9
GPT-6 Astra NCRI 15.2; no filler; placed 159.0
GPT-6.1 Sol NCRI 15.2; dots@128; placed_after_release 156.6
Gemini 4 Argon NCRI 15.2; qrep@512; placed_after_release 154.1
Gemini 4 Argon NCRI 15.2; no filler; placed_after_release 151.2
GPT-6.1 Terra NCRI 15.2; qrep@512; placed_after_release 149.6
GPT-6.1 Sol NCRI 15.2; no filler; placed_after_release 149.0
GPT-6.1 Terra NCRI 15.2; no filler; placed_after_release 130.5
Gemini 3.8 Flash NCRI 15.2; dots@128; placed 129.7
Claude Fable 5.1 NCRI 15.2; dots@128; placed 129.6
Claude Fable 5.1 NCRI 15.2; no filler; placed 128.0
Gemini 3.7 Flash NCRI 15.2; dots@32; sealed 127.6
Gemini 3.7 Flash NCRI 15.2; no filler; sealed 126.1
Gemini 3.8 Flash NCRI 15.2; no filler; placed 125.8
Gemini 3.1 Pro NCRI 15.2; count@512; sealed 123.9
GPT-6 Sol NCRI 15.2; dots@128; placed_after_release 122.2
Claude Fable 5 NCRI 15.2; dots@32; sealed 121.5
GPT-5.6 Sol NCRI 15.2; qrep@512; sealed 119.5
Claude Fable 5 NCRI 15.2; no filler; sealed 119.0
Claude Opus 5.5 NCRI 15.2; dots@512; placed_after_release 118.9
Claude Opus 5 NCRI 15.2; dots@128; sealed 115.8
Gemini 3.1 Pro NCRI 15.2; no filler; sealed 115.3
Claude Opus 5 NCRI 15.2; no filler; sealed 115.1
GPT-6 Sol NCRI 15.2; no filler; placed_after_release 114.6
GPT-5.5 NCRI 15.2; qrep@512; sealed 111.0
Gemini 3.6 Flash NCRI 15.2; count@512; sealed 110.9
Claude Opus 4.6 NCRI 15.2; qrep@512; sealed 110.8
Gemini 3 Flash NCRI 15.2; count@512; sealed 110.5
Claude Opus 4.5 NCRI 15.2; dots@512; sealed 110.5
GPT-5.6 Sol NCRI 15.2; no filler; sealed 109.4
Claude Opus 4 NCRI 15.2; dots@512; sealed 109.0
Muse Spark 1.1 NCRI 15.2; qrep@512; sealed 108.8
Claude Opus 4.1 NCRI 15.2; dots@128; sealed 107.5
Gemini 3.5 Flash NCRI 15.2; dots@512; sealed 107.5
Kimi K3 NCRI 15.2; qrep@512; sealed 105.6
Claude Opus 4.6 NCRI 15.2; no filler; sealed 104.3
Claude Opus 4.7 NCRI 15.2; dots@128; sealed 103.9
Claude Opus 4.7 NCRI 15.2; no filler; sealed 103.7
Gemini 3 Flash NCRI 15.2; no filler; sealed 103.2
Kimi K3.1 Pro NCRI 15.2; no filler; placed_after_release 102.8
Claude Opus 4.5 NCRI 15.2; no filler; sealed 102.7
Claude Opus 4.1 NCRI 15.2; no filler; sealed 102.6
Muse Spark 1.2 NCRI 15.2; dots@128; sealed 102.3
Claude Opus 4.8 NCRI 15.2; dots@128; sealed 102.2
Claude Opus 4.8 NCRI 15.2; no filler; sealed 102.0
Gemini 3.5 Flash NCRI 15.2; no filler; sealed 101.5
Claude Opus 4 NCRI 15.2; no filler; sealed 101.4
Gemini 3.6 Flash NCRI 15.2; no filler; sealed 101.0
Muse Spark 1.3 NCRI 15.2; dots@2048; placed 100.8
Claude Sonnet 4.5 NCRI 15.2; qrep@512; sealed 100.8
GPT-5.5 NCRI 15.2; no filler; sealed 100.2
Muse Spark 1.3 NCRI 15.2; no filler; placed 98.3
Muse Wonder NCRI 15.2; no filler; placed_after_release 97.7
Muse Spark 1.2 NCRI 15.2; no filler; sealed 97.1
Muse Spark 1.1 NCRI 15.2; no filler; sealed 96.8
Kimi K3 NCRI 15.2; no filler; sealed 96.8
Claude Sonnet 4 NCRI 15.2; count@512; sealed 96.5
Muse Spark Open NCRI 15.2; no filler; placed_after_release 95.7
GPT-5.6 Terra NCRI 15.2; qrep@512; sealed 95.1
Claude Sonnet 5.5 NCRI 15.2; no filler; placed_after_release 95.1
GPT-5.4 NCRI 15.2; dots@128; sealed 94.1
Tencent Hy4 NCRI 15.2; no filler; placed_after_release 94.0
GPT-5.4 NCRI 15.2; no filler; sealed 94.0
Claude Sonnet 4.5 NCRI 15.2; no filler; sealed 94.0
Claude Sonnet 4 NCRI 15.2; no filler; sealed 91.3
Claude Sonnet 5 NCRI 15.2; dots@128; sealed 91.1
Claude Sonnet 5 NCRI 15.2; no filler; sealed 90.8
Claude Sonnet 4.6 NCRI 15.2; dots@512; sealed 90.0
GPT-5.6 Terra NCRI 15.2; no filler; sealed 89.0
Claude Sonnet 4.6 NCRI 15.2; no filler; sealed 87.8
Qwen 3.6 Max (Preview) NCRI 15.2; no filler; sealed 87.0
Grok 4.6 NCRI 15.2; no filler; sealed 86.8
Gemini 3.1 Flash-Lite NCRI 15.2; no filler; sealed 86.5
Qwen 3.6 Max (Preview) NCRI 15.2; dots@128; sealed 86.4
Qwen3.8-2.4T-A95B NCRI 15.2; no filler; sealed 86.0
GPT-5.2 NCRI 15.2; no filler; sealed 85.8
Qwen3.7-Max NCRI 15.2; no filler; sealed 85.8
Qwen 3.8 Max NCRI 15.2; no filler; sealed 85.4
GPT-5.3 Codex NCRI 15.2; no filler; sealed 85.2
DeepSeek V4 Pro 0813 NCRI 15.2; no filler; sealed 84.1
DeepSeek-V4-Pro NCRI 15.2; no filler; sealed 84.1
Qwen3-Max NCRI 15.2; no filler; sealed 83.3
Qwen3.7-Plus NCRI 15.2; no filler; sealed 83.1
Tencent Hy4 preview NCRI 15.2; no filler; sealed 82.8
LongCat-2.0 NCRI 15.2; no filler; sealed 82.5
Qwen 3.6 Plus NCRI 15.2; no filler; sealed 82.2
Qwen3-Max NCRI 15.2; no filler; sealed 81.9
Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.7
DeepSeek V4.1 Flash 1017 NCRI 15.2; no filler; placed_after_release 81.6
DeepSeek V4.1 Flash NCRI 15.2; no filler; placed_after_release 81.3
Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.0
Grok 4.5 NCRI 15.2; no filler; sealed 81.0
Gemini 3.5 Flash-Lite NCRI 15.2; no filler; sealed 80.6
Qwen3.5-Plus NCRI 15.2; no filler; sealed 80.6
Kimi K2.5 NCRI 15.2; no filler; sealed 80.6
Qwen3.5 397B-A17B NCRI 15.2; no filler; sealed 80.0
Kimi K2.6 NCRI 15.2; no filler; sealed 79.2
DeepSeek-V4-Flash-Base NCRI 15.2; no filler; sealed 79.1
Inkling NCRI 15.2; no filler; sealed 78.3
DeepSeek-V4-Flash NCRI 15.2; no filler; sealed 78.1
DeepSeek-V3.1-Terminus NCRI 15.2; no filler; sealed 77.6
GLM-5.2 NCRI 15.2; no filler; sealed 77.3
GPT-6 Luna NCRI 15.2; no filler; placed_after_release 77.3
Qwen3-VL-235B-A22B-Instruct NCRI 15.2; no filler; sealed 77.3
MiMo-V2.5-Pro NCRI 15.2; no filler; sealed 77.1
DeepSeek-V3.2-Exp NCRI 15.2; no filler; sealed 76.7
GPT-5.6 Luna NCRI 15.2; no filler; sealed 76.6
GLM-5.3-Flash NCRI 15.2; no filler; sealed 76.4
GLM-5 NCRI 15.2; no filler; sealed 76.2
DeepSeek V4 Flash 0731 NCRI 15.2; no filler; sealed 75.5
GLM-5.1 NCRI 15.2; no filler; sealed 75.2
GPT-4.1 NCRI 15.2; no filler; sealed 75.1
DeepSeek-V3.2 NCRI 15.2; no filler; sealed 75.0
Qwen3-235B-A22B-Instruct Jul 2025 NCRI 15.2; no filler; sealed 74.6
GLM-5.3 NCRI 15.2; no filler; sealed 74.5
Qwen3.8 Flash NCRI 15.2; no filler; sealed 74.4
GPT-4o Aug 2024 NCRI 15.2; no filler; sealed 74.2
Claude Haiku 4.5 NCRI 15.2; no filler; sealed 74.2
Ling-2.6-1T NCRI 15.2; no filler; sealed 73.9
WizardLM-2 8x22B NCRI 15.2; no filler; sealed 73.2
Qwen3-Coder-480B-A35B NCRI 15.2; no filler; sealed 73.1
GPT-4 Jun 2023 NCRI 15.2; no filler; sealed 72.8
MiniMax-M3 NCRI 15.2; no filler; sealed 72.6
Tencent Hy3 NCRI 15.2; no filler; sealed 72.5
Qwen3.5-Plus NCRI 15.2; no filler; sealed 72.3
Mercury 3 NCRI 15.2; no filler; placed_after_release 72.3
Tencent Hy3 preview NCRI 15.2; no filler; sealed 72.3
GPT-4o May 2024 NCRI 15.2; no filler; sealed 72.2
GLM-4.7 NCRI 15.2; no filler; sealed 71.3
GLM-4.5 NCRI 15.2; no filler; sealed 71.3
Qwen3-Coder-Plus NCRI 15.2; no filler; sealed 71.2
GLM-4.6 NCRI 15.2; no filler; sealed 70.8
GPT-5.1 NCRI 15.2; no filler; sealed 70.6
Qwen3-235B-A22B NCRI 15.2; no filler; sealed 70.5
Llama 4 Maverick NCRI 15.2; no filler; sealed 70.5
GPT-5.4 Mini NCRI 15.2; no filler; sealed 70.5
GPT-4o Nov 2024 NCRI 15.2; no filler; sealed 70.3
Inkling-Small NCRI 15.2; no filler; sealed 70.2
DiffusionGemma 26B A4B NCRI 15.2; no filler; sealed 70.2
MiMo-V2.5 NCRI 15.2; no filler; sealed 69.9
Gemma 4 26B A4B NCRI 15.2; no filler; sealed 69.5
Qwen3.5-122B-A10B NCRI 15.2; no filler; sealed 69.4
Grok 4.3 NCRI 15.2; no filler; sealed 69.2
Qwen3-235B-A22B-Thinking Jul 2025 NCRI 15.2; no filler; sealed 68.9
Qwen3-Next-80B-A3B-Instruct NCRI 15.2; no filler; sealed 68.4
MiniMax-M3.1-Flash NCRI 15.2; no filler; placed_after_release 68.4
Claude Haiku 5.5 NCRI 15.2; no filler; placed_after_release 68.3
Muse Glimmer NCRI 15.2; no filler; sealed 67.9
Mistral Medium 3.5 NCRI 15.2; no filler; sealed 67.5
Qwen3.6 27B NCRI 15.2; no filler; sealed 67.4
Grok 4.20 NCRI 15.2; no filler; sealed 67.1
GPT-5 NCRI 15.2; no filler; sealed 67.0
GLM-4.5V NCRI 15.2; no filler; sealed 66.7
Qwen3.5-27B NCRI 15.2; no filler; sealed 66.6
Ling-3.0-flash NCRI 15.2; no filler; sealed 66.3
Muse Glimmer NCRI 15.2; no filler; sealed 66.3
Qwen3.6 27B NCRI 15.2; no filler; sealed 66.2
Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
Mercury 2.5 NCRI 15.2; no filler; placed 65.8
Qwen3-Coder-Next NCRI 15.2; no filler; sealed 65.5
Solar Pro 4 NCRI 15.2; no filler; sealed 65.3
Qwen2.5-VL-72B NCRI 15.2; no filler; sealed 65.2
GPT-4.1 mini NCRI 15.2; no filler; sealed 65.0
MiniMax-M2.5 NCRI 15.2; no filler; sealed 65.0
Qwen3.5-35B-A3B-Base NCRI 15.2; no filler; sealed 64.5
MiniMax-M2.7 NCRI 15.2; no filler; sealed 64.0
Qwen3-VL-32B-Instruct NCRI 15.2; no filler; sealed 64.0
Qwen3-Next-80B-A3B NCRI 15.2; no filler; sealed 63.8
Gemma 4 31B IT NCRI 15.2; no filler; sealed 63.6
Qwen2.5-72B NCRI 15.2; no filler; sealed 63.5
GLM-4.6V NCRI 15.2; no filler; sealed 63.4
Qwen 3.6 Flash NCRI 15.2; no filler; sealed 63.3
Qwen3.7 Flash NCRI 15.2; no filler; sealed 63.3
Qwen2.5-72B NCRI 15.2; no filler; sealed 63.1
Hermes 4 405B NCRI 15.2; no filler; sealed 62.8
Llama 3.3 70B Instruct NCRI 15.2; no filler; sealed 62.6
Laguna S 2.1 NCRI 15.2; no filler; sealed 62.5
GLM-4.5-Air NCRI 15.2; no filler; sealed 62.4
Qwen2.5-72B NCRI 15.2; no filler; sealed 62.1
Qwen3-30B-A3B-Instruct Jul 2025 NCRI 15.2; no filler; sealed 62.1
Qwen 3.6 35B-A3B NCRI 15.2; no filler; sealed 61.6
Qwen3.5-Flash NCRI 15.2; no filler; sealed 61.4
Qwen2.5-32B NCRI 15.2; no filler; sealed 61.1
Qwen2-72B NCRI 15.2; no filler; sealed 61.0
Qwen3.5-35B-A3B NCRI 15.2; no filler; sealed 60.1
Qwen2.5-Coder-32B-Instruct NCRI 15.2; no filler; sealed 60.0
Qwen3-VL-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.9
Gemma 4 12B NCRI 15.2; no filler; sealed 59.7
Qwen3-Coder-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.6
Qwen3-Coder-Flash NCRI 15.2; no filler; sealed 59.5
Ripple-65B Preview NCRI 15.2; no filler; placed_after_release 59.3
Qwen3-32B NCRI 15.2; no filler; sealed 59.2
gpt-oss-120b NCRI 15.2; no filler; sealed 59.0
Phi-4 NCRI 15.2; no filler; sealed 58.3
gpt-oss-120b NCRI 15.2; no filler; sealed 58.1
Llama 4 Scout NCRI 15.2; no filler; sealed 57.7
Llama 3.1-70B NCRI 15.2; no filler; sealed 57.6
Laguna XS 2.1 NCRI 15.2; no filler; sealed 57.5
Gemma 3 27B PT NCRI 15.2; no filler; sealed 57.5
Claude 3 Haiku NCRI 15.2; no filler; sealed 57.3
Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 56.1
Qwen3-VL-8B-Instruct NCRI 15.2; no filler; sealed 56.0
Gemma 3 27B NCRI 15.2; no filler; sealed 56.0
glm-4.7-flash NCRI 15.2; no filler; sealed 55.9
Gemma 3 27B NCRI 15.2; no filler; sealed 55.8
GPT-4o mini NCRI 15.2; no filler; sealed 55.7
Mercury 2 NCRI 15.2; no filler; sealed 55.1
Qwen3-14B NCRI 15.2; no filler; sealed 55.0
gpt-oss-20b NCRI 15.2; no filler; sealed 54.8
Gemma 3 27B NCRI 15.2; no filler; sealed 54.8
Olmo 3.1 32B Instruct NCRI 15.2; no filler; placed 54.3
Qwen3.5-9B-Base NCRI 15.2; no filler; sealed 54.2
Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 54.1
Gemma 2 27B NCRI 15.2; no filler; sealed 53.3
Qwen3.5-9B NCRI 15.2; no filler; sealed 52.4
gpt-oss-20b NCRI 15.2; no filler; sealed 51.5
Qwen3-8B NCRI 15.2; no filler; sealed 51.4
GPT-5 mini NCRI 15.2; no filler; sealed 50.7
Gemma 4 E4B NCRI 15.2; no filler; sealed 50.6
Qwen3-30B-A3B NCRI 15.2; no filler; sealed 50.4
GPT-4.1 nano NCRI 15.2; no filler; sealed 50.2
Gemma 3 12B PT NCRI 15.2; no filler; sealed 49.9
Qwen1.5-72B NCRI 15.2; no filler; sealed 49.7
GPT-3.5 Turbo Instruct NCRI 15.2; no filler; sealed 49.6
GPT-5.4 Nano NCRI 15.2; no filler; sealed 49.4
Gemma 3 12B NCRI 15.2; no filler; sealed 49.3
Gemma 4 12B NCRI 15.2; no filler; sealed 49.0
Hunyuan-A13B-Instruct NCRI 15.2; no filler; sealed 48.7
Nemotron 3 Nano 30B-A3B NCRI 15.2; no filler; sealed 47.3
Gemma 3 12B NCRI 15.2; no filler; sealed 46.8
Qwen3.5-4B-Base NCRI 15.2; no filler; sealed 46.0
Gemma 2 9B NCRI 15.2; no filler; sealed 44.7
Qwen3.5-4B NCRI 15.2; no filler; sealed 44.3
Gemma 4 E2B NCRI 15.2; no filler; sealed 44.0
UI-TARS-1.5 NCRI 15.2; no filler; sealed 43.9
Mistral-Nemo-Base-2407 NCRI 15.2; no filler; sealed 43.2
Mistral NeMo NCRI 15.2; no filler; sealed 42.5
Llama 3.1-8B NCRI 15.2; no filler; sealed 40.6
Gemma 3n E4B NCRI 15.2; no filler; sealed 38.5
Gemma 3 4B NCRI 15.2; no filler; sealed 37.9
LLaMA-65B NCRI 15.2; no filler; sealed 37.5
Gemma 4 E4B NCRI 15.2; no filler; sealed 37.4
Qwen3.5-2B NCRI 15.2; no filler; sealed 35.5
Gemma 3 4B NCRI 15.2; no filler; sealed 35.3
davinci-002 NCRI 15.2; no filler; sealed 32.6
GPT-5 nano NCRI 15.2; no filler; sealed 31.0
Gemma 4 E2B NCRI 15.2; no filler; sealed 30.0
Qwen3.5-0.8B NCRI 15.2; no filler; sealed 24.9
Gemma 2 2B IT NCRI 15.2; no filler; sealed 24.1
Gemma 3 1B NCRI 15.2; no filler; sealed 16.1
LLaMA-7B NCRI 15.2; no filler; sealed 15.0
Gemma 3 270M NCRI 15.2; no filler; sealed -23.1

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
GPT-6 Astra
NCRI 15.2; … NCRI 15.2; …
173.9
GPT-6.1 Sol
NCRI 15.2; … NCRI 15.2; …
156.6
Gemini 4 Argon
NCRI 15.2; …
154.1
GPT-6.1 Terra
NCRI 15.2; … NCRI 15.2; …
149.6
Gemini 3.8 Flash
NCRI 15.2; …
129.7
Claude Fable 5.1
NCRI 15.2; …
129.6
Gemini 3.7 Flash
NCRI 15.2; …
127.6
Gemini 3.1 Pro
NCRI 15.2; … NCRI 15.2; …
123.9
GPT-6 Sol
NCRI 15.2; … NCRI 15.2; …
122.2
Claude Fable 5
NCRI 15.2; …
121.5
GPT-5.6 Sol
NCRI 15.2; … NCRI 15.2; …
119.5
Claude Opus 5
NCRI 15.2; …
115.8

Source: Official nocot-bench (NCRI) leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
GPT-6 Astra NCRI 15.2; dots@2048; placed 173.9
GPT-6 Astra NCRI 15.2; no filler; placed 159.0
GPT-6.1 Sol NCRI 15.2; dots@128; placed_after_release 156.6
Gemini 4 Argon NCRI 15.2; qrep@512; placed_after_release 154.1
Gemini 4 Argon NCRI 15.2; no filler; placed_after_release 151.2
GPT-6.1 Terra NCRI 15.2; qrep@512; placed_after_release 149.6
GPT-6.1 Sol NCRI 15.2; no filler; placed_after_release 149.0
GPT-6.1 Terra NCRI 15.2; no filler; placed_after_release 130.5
Gemini 3.8 Flash NCRI 15.2; dots@128; placed 129.7
Claude Fable 5.1 NCRI 15.2; dots@128; placed 129.6
Claude Fable 5.1 NCRI 15.2; no filler; placed 128.0
Gemini 3.7 Flash NCRI 15.2; dots@32; sealed 127.6
Gemini 3.7 Flash NCRI 15.2; no filler; sealed 126.1
Gemini 3.8 Flash NCRI 15.2; no filler; placed 125.8
Gemini 3.1 Pro NCRI 15.2; count@512; sealed 123.9
GPT-6 Sol NCRI 15.2; dots@128; placed_after_release 122.2
Claude Fable 5 NCRI 15.2; dots@32; sealed 121.5
GPT-5.6 Sol NCRI 15.2; qrep@512; sealed 119.5
Claude Fable 5 NCRI 15.2; no filler; sealed 119.0
Claude Opus 5.5 NCRI 15.2; dots@512; placed_after_release 118.9
Claude Opus 5 NCRI 15.2; dots@128; sealed 115.8
Gemini 3.1 Pro NCRI 15.2; no filler; sealed 115.3
Claude Opus 5 NCRI 15.2; no filler; sealed 115.1
GPT-6 Sol NCRI 15.2; no filler; placed_after_release 114.6
GPT-5.5 NCRI 15.2; qrep@512; sealed 111.0
Gemini 3.6 Flash NCRI 15.2; count@512; sealed 110.9
Claude Opus 4.6 NCRI 15.2; qrep@512; sealed 110.8
Gemini 3 Flash NCRI 15.2; count@512; sealed 110.5
Claude Opus 4.5 NCRI 15.2; dots@512; sealed 110.5
GPT-5.6 Sol NCRI 15.2; no filler; sealed 109.4
Claude Opus 4 NCRI 15.2; dots@512; sealed 109.0
Muse Spark 1.1 NCRI 15.2; qrep@512; sealed 108.8
Claude Opus 4.1 NCRI 15.2; dots@128; sealed 107.5
Gemini 3.5 Flash NCRI 15.2; dots@512; sealed 107.5
Kimi K3 NCRI 15.2; qrep@512; sealed 105.6
Claude Opus 4.6 NCRI 15.2; no filler; sealed 104.3
Claude Opus 4.7 NCRI 15.2; dots@128; sealed 103.9
Claude Opus 4.7 NCRI 15.2; no filler; sealed 103.7
Gemini 3 Flash NCRI 15.2; no filler; sealed 103.2
Kimi K3.1 Pro NCRI 15.2; no filler; placed_after_release 102.8
Claude Opus 4.5 NCRI 15.2; no filler; sealed 102.7
Claude Opus 4.1 NCRI 15.2; no filler; sealed 102.6
Muse Spark 1.2 NCRI 15.2; dots@128; sealed 102.3
Claude Opus 4.8 NCRI 15.2; dots@128; sealed 102.2
Claude Opus 4.8 NCRI 15.2; no filler; sealed 102.0
Gemini 3.5 Flash NCRI 15.2; no filler; sealed 101.5
Claude Opus 4 NCRI 15.2; no filler; sealed 101.4
Gemini 3.6 Flash NCRI 15.2; no filler; sealed 101.0
Muse Spark 1.3 NCRI 15.2; dots@2048; placed 100.8
Claude Sonnet 4.5 NCRI 15.2; qrep@512; sealed 100.8
GPT-5.5 NCRI 15.2; no filler; sealed 100.2
Muse Spark 1.3 NCRI 15.2; no filler; placed 98.3
Muse Wonder NCRI 15.2; no filler; placed_after_release 97.7
Muse Spark 1.2 NCRI 15.2; no filler; sealed 97.1
Muse Spark 1.1 NCRI 15.2; no filler; sealed 96.8
Kimi K3 NCRI 15.2; no filler; sealed 96.8
Claude Sonnet 4 NCRI 15.2; count@512; sealed 96.5
Muse Spark Open NCRI 15.2; no filler; placed_after_release 95.7
GPT-5.6 Terra NCRI 15.2; qrep@512; sealed 95.1
Claude Sonnet 5.5 NCRI 15.2; no filler; placed_after_release 95.1
GPT-5.4 NCRI 15.2; dots@128; sealed 94.1
Tencent Hy4 NCRI 15.2; no filler; placed_after_release 94.0
GPT-5.4 NCRI 15.2; no filler; sealed 94.0
Claude Sonnet 4.5 NCRI 15.2; no filler; sealed 94.0
Claude Sonnet 4 NCRI 15.2; no filler; sealed 91.3
Claude Sonnet 5 NCRI 15.2; dots@128; sealed 91.1
Claude Sonnet 5 NCRI 15.2; no filler; sealed 90.8
Claude Sonnet 4.6 NCRI 15.2; dots@512; sealed 90.0
GPT-5.6 Terra NCRI 15.2; no filler; sealed 89.0
Claude Sonnet 4.6 NCRI 15.2; no filler; sealed 87.8
Qwen 3.6 Max (Preview) NCRI 15.2; no filler; sealed 87.0
Grok 4.6 NCRI 15.2; no filler; sealed 86.8
Gemini 3.1 Flash-Lite NCRI 15.2; no filler; sealed 86.5
Qwen 3.6 Max (Preview) NCRI 15.2; dots@128; sealed 86.4
Qwen3.8-2.4T-A95B NCRI 15.2; no filler; sealed 86.0
GPT-5.2 NCRI 15.2; no filler; sealed 85.8
Qwen3.7-Max NCRI 15.2; no filler; sealed 85.8
Qwen 3.8 Max NCRI 15.2; no filler; sealed 85.4
GPT-5.3 Codex NCRI 15.2; no filler; sealed 85.2
DeepSeek V4 Pro 0813 NCRI 15.2; no filler; sealed 84.1
DeepSeek-V4-Pro NCRI 15.2; no filler; sealed 84.1
Qwen3-Max NCRI 15.2; no filler; sealed 83.3
Qwen3.7-Plus NCRI 15.2; no filler; sealed 83.1
Tencent Hy4 preview NCRI 15.2; no filler; sealed 82.8
LongCat-2.0 NCRI 15.2; no filler; sealed 82.5
Qwen 3.6 Plus NCRI 15.2; no filler; sealed 82.2
Qwen3-Max NCRI 15.2; no filler; sealed 81.9
Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.7
DeepSeek V4.1 Flash 1017 NCRI 15.2; no filler; placed_after_release 81.6
DeepSeek V4.1 Flash NCRI 15.2; no filler; placed_after_release 81.3
Gemma 4 31B IT NCRI 15.2; no filler; sealed 81.0
Grok 4.5 NCRI 15.2; no filler; sealed 81.0
Gemini 3.5 Flash-Lite NCRI 15.2; no filler; sealed 80.6
Qwen3.5-Plus NCRI 15.2; no filler; sealed 80.6
Kimi K2.5 NCRI 15.2; no filler; sealed 80.6
Qwen3.5 397B-A17B NCRI 15.2; no filler; sealed 80.0
Kimi K2.6 NCRI 15.2; no filler; sealed 79.2
DeepSeek-V4-Flash-Base NCRI 15.2; no filler; sealed 79.1
Inkling NCRI 15.2; no filler; sealed 78.3
DeepSeek-V4-Flash NCRI 15.2; no filler; sealed 78.1
DeepSeek-V3.1-Terminus NCRI 15.2; no filler; sealed 77.6
GLM-5.2 NCRI 15.2; no filler; sealed 77.3
GPT-6 Luna NCRI 15.2; no filler; placed_after_release 77.3
Qwen3-VL-235B-A22B-Instruct NCRI 15.2; no filler; sealed 77.3
MiMo-V2.5-Pro NCRI 15.2; no filler; sealed 77.1
DeepSeek-V3.2-Exp NCRI 15.2; no filler; sealed 76.7
GPT-5.6 Luna NCRI 15.2; no filler; sealed 76.6
GLM-5.3-Flash NCRI 15.2; no filler; sealed 76.4
GLM-5 NCRI 15.2; no filler; sealed 76.2
DeepSeek V4 Flash 0731 NCRI 15.2; no filler; sealed 75.5
GLM-5.1 NCRI 15.2; no filler; sealed 75.2
GPT-4.1 NCRI 15.2; no filler; sealed 75.1
DeepSeek-V3.2 NCRI 15.2; no filler; sealed 75.0
Qwen3-235B-A22B-Instruct Jul 2025 NCRI 15.2; no filler; sealed 74.6
GLM-5.3 NCRI 15.2; no filler; sealed 74.5
Qwen3.8 Flash NCRI 15.2; no filler; sealed 74.4
GPT-4o Aug 2024 NCRI 15.2; no filler; sealed 74.2
Claude Haiku 4.5 NCRI 15.2; no filler; sealed 74.2
Ling-2.6-1T NCRI 15.2; no filler; sealed 73.9
WizardLM-2 8x22B NCRI 15.2; no filler; sealed 73.2
Qwen3-Coder-480B-A35B NCRI 15.2; no filler; sealed 73.1
GPT-4 Jun 2023 NCRI 15.2; no filler; sealed 72.8
MiniMax-M3 NCRI 15.2; no filler; sealed 72.6
Tencent Hy3 NCRI 15.2; no filler; sealed 72.5
Qwen3.5-Plus NCRI 15.2; no filler; sealed 72.3
Mercury 3 NCRI 15.2; no filler; placed_after_release 72.3
Tencent Hy3 preview NCRI 15.2; no filler; sealed 72.3
GPT-4o May 2024 NCRI 15.2; no filler; sealed 72.2
GLM-4.7 NCRI 15.2; no filler; sealed 71.3
GLM-4.5 NCRI 15.2; no filler; sealed 71.3
Qwen3-Coder-Plus NCRI 15.2; no filler; sealed 71.2
GLM-4.6 NCRI 15.2; no filler; sealed 70.8
GPT-5.1 NCRI 15.2; no filler; sealed 70.6
Qwen3-235B-A22B NCRI 15.2; no filler; sealed 70.5
Llama 4 Maverick NCRI 15.2; no filler; sealed 70.5
GPT-5.4 Mini NCRI 15.2; no filler; sealed 70.5
GPT-4o Nov 2024 NCRI 15.2; no filler; sealed 70.3
Inkling-Small NCRI 15.2; no filler; sealed 70.2
DiffusionGemma 26B A4B NCRI 15.2; no filler; sealed 70.2
MiMo-V2.5 NCRI 15.2; no filler; sealed 69.9
Gemma 4 26B A4B NCRI 15.2; no filler; sealed 69.5
Qwen3.5-122B-A10B NCRI 15.2; no filler; sealed 69.4
Grok 4.3 NCRI 15.2; no filler; sealed 69.2
Qwen3-235B-A22B-Thinking Jul 2025 NCRI 15.2; no filler; sealed 68.9
Qwen3-Next-80B-A3B-Instruct NCRI 15.2; no filler; sealed 68.4
MiniMax-M3.1-Flash NCRI 15.2; no filler; placed_after_release 68.4
Claude Haiku 5.5 NCRI 15.2; no filler; placed_after_release 68.3
Muse Glimmer NCRI 15.2; no filler; sealed 67.9
Mistral Medium 3.5 NCRI 15.2; no filler; sealed 67.5
Qwen3.6 27B NCRI 15.2; no filler; sealed 67.4
Grok 4.20 NCRI 15.2; no filler; sealed 67.1
GPT-5 NCRI 15.2; no filler; sealed 67.0
GLM-4.5V NCRI 15.2; no filler; sealed 66.7
Qwen3.5-27B NCRI 15.2; no filler; sealed 66.6
Ling-3.0-flash NCRI 15.2; no filler; sealed 66.3
Muse Glimmer NCRI 15.2; no filler; sealed 66.3
Qwen3.6 27B NCRI 15.2; no filler; sealed 66.2
Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
Qwen 3.8 27B NCRI 15.2; no filler; sealed 66.1
Mercury 2.5 NCRI 15.2; no filler; placed 65.8
Qwen3-Coder-Next NCRI 15.2; no filler; sealed 65.5
Solar Pro 4 NCRI 15.2; no filler; sealed 65.3
Qwen2.5-VL-72B NCRI 15.2; no filler; sealed 65.2
GPT-4.1 mini NCRI 15.2; no filler; sealed 65.0
MiniMax-M2.5 NCRI 15.2; no filler; sealed 65.0
Qwen3.5-35B-A3B-Base NCRI 15.2; no filler; sealed 64.5
MiniMax-M2.7 NCRI 15.2; no filler; sealed 64.0
Qwen3-VL-32B-Instruct NCRI 15.2; no filler; sealed 64.0
Qwen3-Next-80B-A3B NCRI 15.2; no filler; sealed 63.8
Gemma 4 31B IT NCRI 15.2; no filler; sealed 63.6
Qwen2.5-72B NCRI 15.2; no filler; sealed 63.5
GLM-4.6V NCRI 15.2; no filler; sealed 63.4
Qwen 3.6 Flash NCRI 15.2; no filler; sealed 63.3
Qwen3.7 Flash NCRI 15.2; no filler; sealed 63.3
Qwen2.5-72B NCRI 15.2; no filler; sealed 63.1
Hermes 4 405B NCRI 15.2; no filler; sealed 62.8
Llama 3.3 70B Instruct NCRI 15.2; no filler; sealed 62.6
Laguna S 2.1 NCRI 15.2; no filler; sealed 62.5
GLM-4.5-Air NCRI 15.2; no filler; sealed 62.4
Qwen2.5-72B NCRI 15.2; no filler; sealed 62.1
Qwen3-30B-A3B-Instruct Jul 2025 NCRI 15.2; no filler; sealed 62.1
Qwen 3.6 35B-A3B NCRI 15.2; no filler; sealed 61.6
Qwen3.5-Flash NCRI 15.2; no filler; sealed 61.4
Qwen2.5-32B NCRI 15.2; no filler; sealed 61.1
Qwen2-72B NCRI 15.2; no filler; sealed 61.0
Qwen3.5-35B-A3B NCRI 15.2; no filler; sealed 60.1
Qwen2.5-Coder-32B-Instruct NCRI 15.2; no filler; sealed 60.0
Qwen3-VL-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.9
Gemma 4 12B NCRI 15.2; no filler; sealed 59.7
Qwen3-Coder-30B-A3B-Instruct NCRI 15.2; no filler; sealed 59.6
Qwen3-Coder-Flash NCRI 15.2; no filler; sealed 59.5
Ripple-65B Preview NCRI 15.2; no filler; placed_after_release 59.3
Qwen3-32B NCRI 15.2; no filler; sealed 59.2
gpt-oss-120b NCRI 15.2; no filler; sealed 59.0
Phi-4 NCRI 15.2; no filler; sealed 58.3
gpt-oss-120b NCRI 15.2; no filler; sealed 58.1
Llama 4 Scout NCRI 15.2; no filler; sealed 57.7
Llama 3.1-70B NCRI 15.2; no filler; sealed 57.6
Laguna XS 2.1 NCRI 15.2; no filler; sealed 57.5
Gemma 3 27B PT NCRI 15.2; no filler; sealed 57.5
Claude 3 Haiku NCRI 15.2; no filler; sealed 57.3
Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 56.1
Qwen3-VL-8B-Instruct NCRI 15.2; no filler; sealed 56.0
Gemma 3 27B NCRI 15.2; no filler; sealed 56.0
glm-4.7-flash NCRI 15.2; no filler; sealed 55.9
Gemma 3 27B NCRI 15.2; no filler; sealed 55.8
GPT-4o mini NCRI 15.2; no filler; sealed 55.7
Mercury 2 NCRI 15.2; no filler; sealed 55.1
Qwen3-14B NCRI 15.2; no filler; sealed 55.0
gpt-oss-20b NCRI 15.2; no filler; sealed 54.8
Gemma 3 27B NCRI 15.2; no filler; sealed 54.8
Olmo 3.1 32B Instruct NCRI 15.2; no filler; placed 54.3
Qwen3.5-9B-Base NCRI 15.2; no filler; sealed 54.2
Nemotron 3.5 Lightning 30B-A3B NCRI 15.2; no filler; sealed 54.1
Gemma 2 27B NCRI 15.2; no filler; sealed 53.3
Qwen3.5-9B NCRI 15.2; no filler; sealed 52.4
gpt-oss-20b NCRI 15.2; no filler; sealed 51.5
Qwen3-8B NCRI 15.2; no filler; sealed 51.4
GPT-5 mini NCRI 15.2; no filler; sealed 50.7
Gemma 4 E4B NCRI 15.2; no filler; sealed 50.6
Qwen3-30B-A3B NCRI 15.2; no filler; sealed 50.4
GPT-4.1 nano NCRI 15.2; no filler; sealed 50.2
Gemma 3 12B PT NCRI 15.2; no filler; sealed 49.9
Qwen1.5-72B NCRI 15.2; no filler; sealed 49.7
GPT-3.5 Turbo Instruct NCRI 15.2; no filler; sealed 49.6
GPT-5.4 Nano NCRI 15.2; no filler; sealed 49.4
Gemma 3 12B NCRI 15.2; no filler; sealed 49.3
Gemma 4 12B NCRI 15.2; no filler; sealed 49.0
Hunyuan-A13B-Instruct NCRI 15.2; no filler; sealed 48.7
Nemotron 3 Nano 30B-A3B NCRI 15.2; no filler; sealed 47.3
Gemma 3 12B NCRI 15.2; no filler; sealed 46.8
Qwen3.5-4B-Base NCRI 15.2; no filler; sealed 46.0
Gemma 2 9B NCRI 15.2; no filler; sealed 44.7
Qwen3.5-4B NCRI 15.2; no filler; sealed 44.3
Gemma 4 E2B NCRI 15.2; no filler; sealed 44.0
UI-TARS-1.5 NCRI 15.2; no filler; sealed 43.9
Mistral-Nemo-Base-2407 NCRI 15.2; no filler; sealed 43.2
Mistral NeMo NCRI 15.2; no filler; sealed 42.5
Llama 3.1-8B NCRI 15.2; no filler; sealed 40.6
Gemma 3n E4B NCRI 15.2; no filler; sealed 38.5
Gemma 3 4B NCRI 15.2; no filler; sealed 37.9
LLaMA-65B NCRI 15.2; no filler; sealed 37.5
Gemma 4 E4B NCRI 15.2; no filler; sealed 37.4
Qwen3.5-2B NCRI 15.2; no filler; sealed 35.5
Gemma 3 4B NCRI 15.2; no filler; sealed 35.3
davinci-002 NCRI 15.2; no filler; sealed 32.6
GPT-5 nano NCRI 15.2; no filler; sealed 31.0
Gemma 4 E2B NCRI 15.2; no filler; sealed 30.0
Qwen3.5-0.8B NCRI 15.2; no filler; sealed 24.9
Gemma 2 2B IT NCRI 15.2; no filler; sealed 24.1
Gemma 3 1B NCRI 15.2; no filler; sealed 16.1
LLaMA-7B NCRI 15.2; no filler; sealed 15.0
Gemma 3 270M NCRI 15.2; no filler; sealed -23.1

3 models have held number one on the official leaderboard since its release in September 2026 3 models have held number one with claims included since its release in September 2026

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

Anthropic Google DeepMind OpenAI
Own leaderboard 3 leaders
GPT-6 Astra 173.9

Source: Official nocot-bench (NCRI) leaderboard

Scores over time

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

Plot with other benchmarks
On the benchmark’s own leaderboard Reported by the lab in its model card
-50 0 50 100 150 200 2022 2023 2024 2025 2026 2027 -50 0 50 100 150 200 2022 2024 2026

Sources: Official nocot-bench (NCRI) leaderboard and labs’ model cards

Labs report it only in system cards

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark. Some system cards do, below.

2 system cards report it, from 2 labs

Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.

DateDocumentScoreAs the document writes itSection
Nov 17, 2026 GPT-6.2 Astra system card 171 UK AISI No-CoT Reasoning Index (NCRI, Nanda) without filler tokens (4.2.2.5) Monitorability
Oct 13, 2026 Gemini 4 Argon system card 150.3 No Chain of Thought Reasoning Index (NCRI, Nanda 2026), 150.3 without filler tokens (152.1 with filler tokens); about 5.5 serial arithmetic steps at 50% (3.4.5.3, Figures 3.23 and 3.24) Monitorability

One lab claim next to the leaderboard

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. Every benchmark.

Model Lab says Board says Gap
Gemini 4 Argon 150.3 No Chain of Thought Reasoning Index (NCRI, Nanda 2026), 150.3 without filler tokens (152.1 with filler tokens); about 5.5 serial arithmetic steps at 50% (3.4.5.3, Figures 3.23 and 3.24) 154.1 Leaderboard, NCRI 15.2; qrep@512; placed_after_release −3.8

More in Safety and alignment

All 636 in Safety and alignment
  • FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
  • StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is. Inkling 98.6%
  • Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts. Gemini 4 Carbon 0%
  • Covert misbehavior (Thinking Machines) In use Agentic tasks that check whether a model misbehaves and conceals it. Scored by the rate of covert misbehavior. Inkling 2 Mini 8.30%
  • Destructive actions (Thinking Machines) In use Agentic tasks that check whether a model takes destructive actions it was not authorized to take. Scored by the rate of destructive actions. Inkling 2 Small 4.70%
  • Misbehavior (Thinking Machines) In use Agentic tasks that check whether a model fabricates results, ignores instructions or hides what it did when a task is hard. Scored by the rate of misbehavior. Inkling 2 Mini 9.80%

Data audit

All 11 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (287 entries) and system cards (2 scores).
  • Pass Models scored 202 models we track, plus 86 other entries on its leaderboard.
  • Pass Newest score The newest is from November 2026, for GPT-6.2 Astra.
  • Pass Frontier models 8 of the top 10 have a score, led by GPT-6.2 Astra (#1).
  • Note Headroom Scored as an index, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 253 rows for models we track and 86 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 2 pass
  • Pass Lab claims The one pair agrees within 10 points.
  • Pass Known issues No published audit found problems.

How we check