Benchmark Almanac

SEND Pedagogy Benchmark

Claude Fable 5.1 89.5% leads since September 2026
Not in model cards

Models choose how a teacher should support students with disabilities or special educational needs. The questions come from Chilean teacher exams.

Fab AI publishes overall accuracy for 275 model configurations. The paper scores 220 questions and uses three more as examples. The dataset requires acceptance of access conditions.

Made by
Fab AI
Released
March 2025
Kind
Exam questions
Level
University
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Subject
Education
Language
Spanish
In and out
Text → Text
Interaction
Single turn
Items from
Exams and contests
Refresh
One fixed set
Run by
Anyone
Occupations
Educational instruction and library
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
October 2026
Scores refreshed
November 20th, 2026
247
Entries on its own leaderboard, 230 models we track

2 models share the top score of 89.5% 2 models share the top score of 89.5%

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.

247 entries. Scroll sideways for the rest.

  1. 89.5 Claude Fable 5.1
  2. 89.5 Gemini 4 Argon
  3. 89.0 Claude Opus 5
  4. 88.5 Gemini 3.6 Flash
  5. 88.1 GPT-5.5
  6. 88.1 Qwen 3.8 Max
  7. 88.1 Muse Spark 1.2
  8. 88.1 GPT-6 Astra
  9. 87.6 GPT-5.6 Sol
  10. 87.6 GPT-6.1 Terra
  11. 87.2 Claude Fable 5
  12. 87.2 Gemini 3.7 Flash
  13. 87.2 Qwen3.8 Max (0902)
  14. 86.7 Gemini 3.1 Pro
  15. 86.7 Muse Wonder
  16. 86.2 Gemini 3.5 Flash
  17. 86.2 Grok 4.6
  18. 86.2 Qwen3.8-2.4T-A95B
  19. 86.2 Gemini 3.8 Flash
  20. 86.2 GPT-6.1 Sol
  21. 85.8 Gemini 2.5 Pro
  22. 85.8 Gemini 3 Flash
  23. 85.3 Claude Opus 4.7
  24. 85.3 Qwen3.7-Max
  25. 85.3 Claude Opus 5.5
  26. 84.9 Grok 4.5
  27. 84.9 Muse Spark 1.3
  28. 84.4 Gemini 3 Pro
  29. 84.4 Qwen3.8 Flash
  30. 84.4 Grok 4.7
  31. 84.4 Claude Sonnet 5.5
  32. 83.9 Grok 4.20
  33. 83.9 GPT-5.4
  34. 83.9 Kimi K3.1 Pro
  35. 83.5 Claude Opus 4
  36. 83.5 Claude Sonnet 4.5
  37. 83.5 Qwen3.5 397B-A17B
  38. 83.5 Muse Spark 1.1
  39. 83.0 Kimi K2.6
  40. 83.0 DeepSeek-V4-Flash
  41. 83.0 Claude Opus 4.8
  42. 83.0 Inkling
  43. 83.0 Tencent Hy4 preview
  44. 82.6 GLM-5
  45. 82.6 Qwen 3.6 Max (Preview)
  46. 82.6 Claude Sonnet 5
  47. 82.1 o3
  48. 82.1 Grok 4
  49. 82.1 GPT-5.1
  50. 82.1 DeepSeek-V4-Pro
  51. 82.1 Grok 4.3
  52. 82.1 GPT-5.6 Terra
  53. 82.1 Inkling-Small
  54. 82.1 GLM-5.3
  55. 81.7 GPT-5
  56. 81.7 GLM-5.1
  57. 81.7 Kimi K3
  58. 81.7 DeepSeek V4.1 Flash
  59. 81.2 Qwen3-235B-A22B-Thinking
  60. 81.2 GLM-4.5
  61. 81.2 Kimi K2 Thinking
  62. 81.2 Qwen3.5-122B-A10B
  63. 81.2 GLM-5.2
  64. 80.7 Claude Sonnet 4
  65. 80.7 Kimi K2
  66. 80.7 Claude Opus 4.5
  67. 80.7 GPT-5.2
  68. 80.7 Claude Opus 4.6
  69. 80.7 Qwen3.5-Plus
  70. 80.7 Claude Sonnet 4.6
  71. 80.7 Qwen 3.6 Plus
  72. 80.7 Gemma 4 31B IT
  73. 80.7 Nemotron 3 Ultra
  74. 80.3 Gemini 2.5 Flash
  75. 80.3 Grok 4.1 Fast
  76. 80.3 GLM-4.7
  77. 80.3 GPT-5.6 Luna
  78. 80.3 DeepSeek V4.1 Flash 1017
  79. 79.8 Gemini 3.1 Flash-Lite
  80. 79.4 o1
  81. 79.4 GPT-4.1
  82. 79.4 GPT-5 mini
  83. 79.4 Mistral Large 3
  84. 79.4 MiMo-V2-Pro
  85. 79.4 MiniMax-M3
  86. 79.4 Muse Glimmer
  87. 79.4 Qwen 3.8 27B
  88. 79.4 Claude Haiku 5.5
  89. 78.9 Claude 3.7 Sonnet
  90. 78.9 Mistral Medium 3.1
  91. 78.9 GLM-4.6
  92. 78.9 Qwen3.5-27B
  93. 78.9 MiMo-V2.6 Pro
  94. 78.4 o4-mini
  95. 78.4 DeepSeek-R1
  96. 78.4 Kimi K2.5
  97. 78.4 MiMo-V2.6 Flash
  98. 78.0 GPT-4.5
  99. 78.0 Qwen3.5-35B-A3B
  100. 78.0 GPT-5.4 Mini
  101. 78.0 Qwen3.6 27B
  102. 78.0 Mistral Large 4
  103. 77.5 Grok 4 Fast
  104. 77.5 Qwen3.5-Flash
  105. 77.5 MiniMax-M2.7
  106. 77.5 Qwen 3.6 35B-A3B
  107. 77.5 MiMo-V2.5
  108. 77.5 Qwen3.7 Flash
  109. 77.1 Grok-2
  110. 77.1 Mistral Medium 3
  111. 77.1 Gemma 4 26B A4B
  112. 77.1 Gemma 4 12B
  113. 76.6 Qwen 3.6 Flash
  114. 76.6 Mistral Medium 3.5
  115. 76.6 Gemini 3.5 Flash-Lite
  116. 76.2 MiniMax-M2.1
  117. 75.7 Gemini 2.0 Flash
  118. 75.7 Grok-3 mini
  119. 75.7 Claude Haiku 4.5
  120. 75.7 MiMo-V2.5-Pro
  121. 75.2 Llama 4 Maverick
  122. 75.2 GPT-4.1 mini
  123. 75.2 Qwen3-235B-A22B
  124. 75.2 Qwen3-30B-A3B
  125. 75.2 Qwen3-32B
  126. 74.8 Magistral Medium 1.0
  127. 74.3 Nemotron 3.5 Lightning 30B-A3B
  128. 73.8 DeepSeek-V3
  129. 73.8 GPT-4o
  130. 73.8 gpt-oss-120b
  131. 73.4 Gemma 2 27B
  132. 73.4 Gemma 3 27B
  133. 73.4 Gemini 2.5 Flash-Lite
  134. 73.4 Command A+
  135. 72.9 Doubao-1.5-pro
  136. 72.9 LFM2.5-24B-A2B
  137. 72.5 Llama 3.1-405B
  138. 72.5 o3-mini
  139. 72.5 DeepSeek-V3.2
  140. 72.5 Qwen3.5-9B
  141. 72.5 GPT-5.4 Nano
  142. 72.0 Gemini 1.5 Pro
  143. 72.0 Cohere Command A
  144. 72.0 Grok 3
  145. 71.6 Gemini 2.0 Flash-Lite
  146. 71.6 MiniMax-M3.1-Flash
  147. 71.1 learnlm-1.5-pro-experimental
  148. 71.1 Ministral 3 8B
  149. 71.1 GPT-4o mini
  150. 71.1 Mistral Large 2
  151. 71.1 Amazon Nova Pro
  152. 70.6 Qwen2.5-72B
  153. 70.6 Llama 3.2 90B
  154. 70.6 GPT-5 nano
  155. 70.6 Gemma 4 E4B
  156. 70.2 GPT-4 Turbo
  157. 70.2 Llama 3.1-70B
  158. 70.2 Qwen3-14B
  159. 70.2 gpt-oss-20b
  160. 70.2 DeepSeek-V3.1
  161. 69.7 Hunyuan Large
  162. 69.7 Llama 3.3 70B
  163. 69.7 Olmo 3.1 32B Think
  164. 69.3 Phi-4
  165. 69.3 Mistral Small 3.1
  166. 69.3 Qwen3-8B
  167. 69.3 Mistral Small 3.2
  168. 68.8 GPT-4
  169. 68.3 Doubao-1.5 Lite
  170. 68.3 Claude 3 Opus
  171. 68.3 Gemini 1.5 Flash
  172. 68.3 o1-mini
  173. 68.3 Amazon Nova Lite
  174. 68.3 Mistral Small 3
  175. 68.3 glm-4.7-flash
  176. 67.9 Llama 4 Scout
  177. 67.9 Nova 2 Lite
  178. 67.4 Hunyuan Large (Long Context)
  179. 67.4 Jamba 1.5 Large
  180. 67.4 Jamba 1.7 Large
  181. 67.4 Nemotron 3 Nano 30B-A3B
  182. 67.4 MiMo-V2-Flash
  183. 67.4 Granite 4.2 8B
  184. 67.0 Grok Beta
  185. 67.0 Claude 3.5 Haiku
  186. 67.0 Ministral 3 14B
  187. 66.1 Claude 3.5 Sonnet
  188. 66.1 Command R+
  189. 66.1 Claude 3.5 Sonnet (October 2024)
  190. 66.1 Jamba 1.6 Large
  191. 66.1 Qwen3-4B
  192. 65.6 Yi-Large
  193. 65.6 Gemma 3 12B
  194. 65.6 Mistral Small 4
  195. 65.1 Llama 3.2 11B
  196. 65.1 sonar
  197. 64.7 Mixtral 8x7B
  198. 64.7 Llama 3.1-8B
  199. 64.2 GPT-4.1 nano
  200. 63.8 Yi-Lightning
  201. 63.8 Qwen3-1.7B
  202. 63.8 Olmo 3.1 32B Instruct
  203. 63.3 Olmo-3 7B Think
  204. 63.3 Phi-3.5-vision-instruct
  205. 62.4 Mixtral 8x22B
  206. 62.4 OLMo 2 32B Instruct
  207. 62.4 Gemma 3n E4B
  208. 61.9 Mistral NeMo
  209. 61.9 Qwen2.5-7B
  210. 61.0 Phi-3.5-mini
  211. 60.5 SmolLM3-3B
  212. 60.5 Amazon Nova Micro
  213. 60.5 Jamba 1.6 Mini
  214. 60.5 Gemma 3n E2B
  215. 60.1 LFM-3B
  216. 59.6 Ministral 3 3B
  217. 59.6 Ministral 8B
  218. 59.6 Gemma 4 E2B
  219. 59.2 Ministral 3B
  220. 59.2 LFM2-8B-A1B
  221. 59.2 LFM3-1.7B
  222. 58.3 LFM-7B
  223. 57.8 Jamba Instruct
  224. 57.8 LFM2-2.6B
  225. 57.8 Phi-4-multimodal
  226. 57.3 Gemini 1.5 Flash 8B
  227. 56.9 Olmo-2 13B
  228. 56.9 Jamba 1.5 Mini
  229. 56.0 Claude 3 Haiku
  230. 55.0 Gemma 3 4B
  231. 54.1 Jamba 1.7 Mini
  232. 53.2 Command R7B
  233. 52.3 GPT-3.5 Turbo
  234. 52.3 LFM2-1.2B
  235. 51.4 Olmo-3 7B
  236. 47.3 Qwen3-0.6B
  237. 46.3 Olmo-2 7B
  238. 45.4 LFM2.5-1.2B-Thinking
  239. 38.1 Gemma 3 1B
  240. 33.9 LFM2-700M
  241. 31.6 LFM2-350M
  242. 28.4 Llama 3.2 1B
  243. 28.0 SmolLM2-1.7B
  244. 28.0 LFM2.5-1.2B-Instruct
  245. 26.6 OLMo 2 1B Instruct
  246. 26.1 SmolLM2-360M
  247. 24.8 SmolLM2-135M
  1. 1 Claude Fable 5.1 89.5%
  2. 2 Gemini 4 Argon high 89.5%
  3. 3 Claude Opus 5 89%
  4. 4 Gemini 3.6 Flash 88.5%
  5. 5 GPT-5.5 88.1%
  6. 6 Qwen 3.8 Max 88.1%
  7. 7 Muse Spark 1.2 88.1%
  8. 8 GPT-6 Astra high 88.1%
  9. 9 GPT-5.6 Sol 87.6%
  10. 10 GPT-6.1 Terra medium 87.6%
  11. 11 Claude Fable 5 87.2%
  12. 12 Gemini 3.7 Flash 87.2%
  13. 13 Qwen3.8 Max (0902) 87.2%
  14. 14 Gemini 3.1 Pro 86.7%
  15. 15 Muse Wonder high 86.7%
  16. 16 Gemini 3.5 Flash 86.2%
  17. 17 Grok 4.6 86.2%
  18. 18 Qwen3.8-2.4T-A95B 86.2%
  19. 19 Gemini 3.8 Flash 86.2%
  20. 20 GPT-6.1 Sol high 86.2%
  21. 21 Gemini 2.5 Pro Jun 2025 85.8%
  22. 22 Gemini 3 Flash 85.8%
  23. 23 Claude Opus 4.7 85.3%
  24. 24 Qwen3.7-Max 85.3%
  25. 25 Claude Opus 5.5 medium 85.3%
  26. 26 Grok 4.5 84.9%
  27. 27 Muse Spark 1.3 84.9%
  28. 28 Gemini 3 Pro 84.4%
  29. 29 Qwen3.8 Flash 84.4%
  30. 30 Grok 4.7 low 84.4%
  31. 31 Claude Sonnet 5.5 medium 84.4%
  32. 32 Grok 4.20 reasoning 83.9%
  33. 33 GPT-5.4 medium 83.9%
  34. 34 Kimi K3.1 Pro 83.9%
  35. 35 Claude Opus 4 Thinking (low) 83.5%
  36. 36 Claude Sonnet 4.5 thinking 83.5%
  37. 37 Qwen3.5 397B-A17B 83.5%
  38. 38 Muse Spark 1.1 83.5%
  39. 39 Kimi K2.6 83%
  40. 40 DeepSeek-V4-Flash 83%
  41. 41 Claude Opus 4.8 83%
  42. 42 Inkling 83%
  43. 43 Tencent Hy4 preview 83%
  44. 44 GLM-5 82.6%
  45. 45 Qwen 3.6 Max (Preview) 82.6%
  46. 46 Claude Sonnet 5 82.6%
  47. 47 o3 82.1%
  48. 48 Grok 4 82.1%
  49. 49 GPT-5.1 medium 82.1%
  50. 50 DeepSeek-V4-Pro 82.1%
  51. 51 Grok 4.3 82.1%
  52. 52 GPT-5.6 Terra 82.1%
  53. 53 Inkling-Small 82.1%
  54. 54 GLM-5.3 82.1%
  55. 55 GPT-5 medium 81.7%
  56. 56 GLM-5.1 81.7%
  57. 57 Kimi K3 81.7%
  58. 58 DeepSeek V4.1 Flash 81.7%
  59. 59 Qwen3-235B-A22B-Thinking Jul 2025 81.2%
  60. 60 GLM-4.5 81.2%
  61. 61 Kimi K2 Thinking 81.2%
  62. 62 Qwen3.5-122B-A10B 81.2%
  63. 63 GLM-5.2 81.2%
  64. 64 Claude Sonnet 4 Thinking (low) 80.7%
  65. 65 Kimi K2 Sep 2025 80.7%
  66. 66 Claude Opus 4.5 80.7%
  67. 67 GPT-5.2 medium 80.7%
  68. 68 Claude Opus 4.6 80.7%
  69. 69 Qwen3.5-Plus 80.7%
  70. 70 Claude Sonnet 4.6 Adaptive Thinking 80.7%
  71. 71 Qwen 3.6 Plus 80.7%
  72. 72 Gemma 4 31B IT 80.7%
  73. 73 Nemotron 3 Ultra 80.7%
  74. 74 Gemini 2.5 Flash Sep 2025 80.3%
  75. 75 Grok 4.1 Fast 80.3%
  76. 76 GLM-4.7 80.3%
  77. 77 GPT-5.6 Luna 80.3%
  78. 78 DeepSeek V4.1 Flash 1017 80.3%
  79. 79 Gemini 3.1 Flash-Lite 79.8%
  80. 80 o1 medium 79.4%
  81. 81 GPT-4.1 79.4%
  82. 82 GPT-5 mini medium 79.4%
  83. 83 Mistral Large 3 79.4%
  84. 84 MiMo-V2-Pro 79.4%
  85. 85 MiniMax-M3 79.4%
  86. 86 Muse Glimmer 79.4%
  87. 87 Qwen 3.8 27B 79.4%
  88. 88 Claude Haiku 5.5 medium 79.4%
  89. 89 Claude 3.7 Sonnet Thinking (medium) 78.9%
  90. 90 Mistral Medium 3.1 78.9%
  91. 91 GLM-4.6 78.9%
  92. 92 Qwen3.5-27B 78.9%
  93. 93 MiMo-V2.6 Pro 78.9%
  94. 94 o4-mini 78.4%
  95. 95 DeepSeek-R1 May 2025 78.4%
  96. 96 Kimi K2.5 78.4%
  97. 97 MiMo-V2.6 Flash 78.4%
  98. 98 GPT-4.5 78%
  99. 99 Qwen3.5-35B-A3B 78%
  100. 100 GPT-5.4 Mini medium 78%
  101. 101 Qwen3.6 27B 78%
  102. 102 Mistral Large 4 78%
  103. 103 Grok 4 Fast 77.5%
  104. 104 Qwen3.5-Flash 77.5%
  105. 105 MiniMax-M2.7 77.5%
  106. 106 Qwen 3.6 35B-A3B 77.5%
  107. 107 MiMo-V2.5 77.5%
  108. 108 Qwen3.7 Flash 77.5%
  109. 109 Grok-2 Dec 2024 77.1%
  110. 110 Mistral Medium 3 77.1%
  111. 111 Gemma 4 26B A4B 77.1%
  112. 112 Gemma 4 12B 77.1%
  113. 113 Qwen 3.6 Flash 76.6%
  114. 114 Mistral Medium 3.5 76.6%
  115. 115 Gemini 3.5 Flash-Lite 76.6%
  116. 116 MiniMax-M2.1 76.2%
  117. 117 Gemini 2.0 Flash Feb 2025 75.7%
  118. 118 Grok-3 mini 75.7%
  119. 119 Claude Haiku 4.5 75.7%
  120. 120 MiMo-V2.5-Pro 75.7%
  121. 121 Llama 4 Maverick 75.2%
  122. 122 GPT-4.1 mini 75.2%
  123. 123 Qwen3-235B-A22B 75.2%
  124. 124 Qwen3-30B-A3B 75.2%
  125. 125 Qwen3-32B 75.2%
  126. 126 Magistral Medium 1.0 74.8%
  127. 127 Nemotron 3.5 Lightning 30B-A3B 74.3%
  128. 128 DeepSeek-V3 Mar 2025 73.8%
  129. 129 GPT-4o Mar 2025 73.8%
  130. 130 gpt-oss-120b 73.8%
  131. 131 Gemma 2 27B 73.4%
  132. 132 Gemma 3 27B 73.4%
  133. 133 Gemini 2.5 Flash-Lite Sep 2025 73.4%
  134. 134 Command A+ 73.4%
  135. 135 Doubao-1.5-pro 72.9%
  136. 136 LFM2.5-24B-A2B 72.9%
  137. 137 Llama 3.1-405B 72.5%
  138. 138 o3-mini 72.5%
  139. 139 DeepSeek-V3.2 72.5%
  140. 140 Qwen3.5-9B 72.5%
  141. 141 GPT-5.4 Nano medium 72.5%
  142. 142 Gemini 1.5 Pro Sep 2024 72%
  143. 143 Cohere Command A 72%
  144. 144 Grok 3 72%
  145. 145 Gemini 2.0 Flash-Lite 71.6%
  146. 146 MiniMax-M3.1-Flash 71.6%
  147. 147 learnlm-1.5-pro-experimental 71.1%
  148. 148 Ministral 3 8B 71.1%
  149. 149 GPT-4o mini 71.1%
  150. 150 Mistral Large 2 Nov 2024 71.1%
  151. 151 Amazon Nova Pro 71.1%
  152. 152 Qwen2.5-72B 70.6%
  153. 153 Llama 3.2 90B 70.6%
  154. 154 GPT-5 nano medium 70.6%
  155. 155 Gemma 4 E4B 70.6%
  156. 156 GPT-4 Turbo Apr 2024 70.2%
  157. 157 Llama 3.1-70B 70.2%
  158. 158 Qwen3-14B 70.2%
  159. 159 gpt-oss-20b 70.2%
  160. 160 DeepSeek-V3.1 70.2%
  161. 161 Hunyuan Large 69.7%
  162. 162 Llama 3.3 70B 69.7%
  163. 163 Olmo 3.1 32B Think 69.7%
  164. 164 Phi-4 69.3%
  165. 165 Mistral Small 3.1 69.3%
  166. 166 Qwen3-8B 69.3%
  167. 167 Mistral Small 3.2 69.3%
  168. 168 GPT-4 Jun 2023 68.8%
  169. 169 Doubao-1.5 Lite 68.3%
  170. 170 Claude 3 Opus 68.3%
  171. 171 Gemini 1.5 Flash Sep 2024 68.3%
  172. 172 o1-mini 68.3%
  173. 173 Amazon Nova Lite 68.3%
  174. 174 Mistral Small 3 68.3%
  175. 175 glm-4.7-flash 68.3%
  176. 176 Llama 4 Scout 67.9%
  177. 177 Nova 2 Lite 67.9%
  178. 178 Hunyuan Large (Long Context) 67.4%
  179. 179 Jamba 1.5 Large 67.4%
  180. 180 Jamba 1.7 Large 67.4%
  181. 181 Nemotron 3 Nano 30B-A3B 67.4%
  182. 182 MiMo-V2-Flash 67.4%
  183. 183 Granite 4.2 8B 67.4%
  184. 184 Grok Beta 67%
  185. 185 Claude 3.5 Haiku 67%
  186. 186 Ministral 3 14B 67%
  187. 187 Claude 3.5 Sonnet 66.1%
  188. 188 Command R+ 66.1%
  189. 189 Claude 3.5 Sonnet (October 2024) 66.1%
  190. 190 Jamba 1.6 Large 66.1%
  191. 191 Qwen3-4B 66.1%
  192. 192 Yi-Large 65.6%
  193. 193 Gemma 3 12B 65.6%
  194. 194 Mistral Small 4 65.6%
  195. 195 Llama 3.2 11B 65.1%
  196. 196 sonar 65.1%
  197. 197 Mixtral 8x7B 64.7%
  198. 198 Llama 3.1-8B 64.7%
  199. 199 GPT-4.1 nano 64.2%
  200. 200 Yi-Lightning 63.8%
  201. 201 Qwen3-1.7B 63.8%
  202. 202 Olmo 3.1 32B Instruct 63.8%
  203. 203 Olmo-3 7B Think 63.3%
  204. 204 Phi-3.5-vision-instruct 63.3%
  205. 205 Mixtral 8x22B 62.4%
  206. 206 OLMo 2 32B Instruct 62.4%
  207. 207 Gemma 3n E4B 62.4%
  208. 208 Mistral NeMo 61.9%
  209. 209 Qwen2.5-7B 61.9%
  210. 210 Phi-3.5-mini 61%
  211. 211 SmolLM3-3B 60.5%
  212. 212 Amazon Nova Micro 60.5%
  213. 213 Jamba 1.6 Mini 60.5%
  214. 214 Gemma 3n E2B 60.5%
  215. 215 LFM-3B 60.1%
  216. 216 Ministral 3 3B 59.6%
  217. 217 Ministral 8B 59.6%
  218. 218 Gemma 4 E2B 59.6%
  219. 219 Ministral 3B 59.2%
  220. 220 LFM2-8B-A1B 59.2%
  221. 221 LFM3-1.7B 59.2%
  222. 222 LFM-7B 58.3%
  223. 223 Jamba Instruct 57.8%
  224. 224 LFM2-2.6B 57.8%
  225. 225 Phi-4-multimodal 57.8%
  226. 226 Gemini 1.5 Flash 8B 57.3%
  227. 227 Olmo-2 13B 56.9%
  228. 228 Jamba 1.5 Mini 56.9%
  229. 229 Claude 3 Haiku 56%
  230. 230 Gemma 3 4B 55%
  231. 231 Jamba 1.7 Mini 54.1%
  232. 232 Command R7B 53.2%
  233. 233 GPT-3.5 Turbo Jan 2024 52.3%
  234. 234 LFM2-1.2B 52.3%
  235. 235 Olmo-3 7B 51.4%
  236. 236 Qwen3-0.6B 47.3%
  237. 237 Olmo-2 7B 46.3%
  238. 238 LFM2.5-1.2B-Thinking 45.4%
  239. 239 Gemma 3 1B 38.1%
  240. 240 LFM2-700M 33.9%
  241. 241 LFM2-350M 31.6%
  242. 242 Llama 3.2 1B 28.4%
  243. 243 SmolLM2-1.7B 28%
  244. 244 LFM2.5-1.2B-Instruct 28%
  245. 245 OLMo 2 1B Instruct 26.6%
  246. 246 SmolLM2-360M 26.1%
  247. 247 SmolLM2-135M 24.8%

Source: Official SEND Pedagogy Benchmark leaderboard

230 models. Scroll sideways for the rest.

  1. 89.5 Claude Fable 5.1
  2. 89.5 Gemini 4 Argon
  3. 89.0 Claude Opus 5
  4. 88.5 Gemini 3.6 Flash
  5. 88.1 GPT-5.5
  6. 88.1 Qwen 3.8 Max
  7. 88.1 Muse Spark 1.2
  8. 88.1 GPT-6 Astra
  9. 87.6 GPT-5.6 Sol
  10. 87.6 GPT-6.1 Terra
  11. 87.2 Claude Fable 5
  12. 87.2 Gemini 3.7 Flash
  13. 87.2 Qwen3.8 Max (0902)
  14. 86.7 Gemini 3.1 Pro
  15. 86.7 Muse Wonder
  16. 86.2 Gemini 3.5 Flash
  17. 86.2 Grok 4.6
  18. 86.2 Qwen3.8-2.4T-A95B
  19. 86.2 Gemini 3.8 Flash
  20. 86.2 GPT-6.1 Sol
  21. 85.8 Gemini 2.5 Pro
  22. 85.8 Gemini 3 Flash
  23. 85.3 Claude Opus 4.7
  24. 85.3 Qwen3.7-Max
  25. 85.3 Claude Opus 5.5
  26. 84.9 Grok 4.5
  27. 84.9 Muse Spark 1.3
  28. 84.4 Gemini 3 Pro
  29. 84.4 Qwen3.8 Flash
  30. 84.4 Grok 4.7
  31. 84.4 Claude Sonnet 5.5
  32. 83.9 Grok 4.20
  33. 83.9 GPT-5.4
  34. 83.9 Kimi K3.1 Pro
  35. 83.5 Claude Opus 4
  36. 83.5 Claude Sonnet 4.5
  37. 83.5 Qwen3.5 397B-A17B
  38. 83.5 Muse Spark 1.1
  39. 83.0 Kimi K2.6
  40. 83.0 DeepSeek-V4-Flash
  41. 83.0 Claude Opus 4.8
  42. 83.0 Inkling
  43. 83.0 Tencent Hy4 preview
  44. 82.6 GLM-5
  45. 82.6 Qwen 3.6 Max (Preview)
  46. 82.6 Claude Sonnet 5
  47. 82.1 o3
  48. 82.1 Grok 4
  49. 82.1 GPT-5.1
  50. 82.1 DeepSeek-V4-Pro
  51. 82.1 Grok 4.3
  52. 82.1 GPT-5.6 Terra
  53. 82.1 Inkling-Small
  54. 82.1 GLM-5.3
  55. 81.7 GPT-5
  56. 81.7 GLM-5.1
  57. 81.7 Kimi K3
  58. 81.7 DeepSeek V4.1 Flash
  59. 81.2 Qwen3-235B-A22B-Thinking
  60. 81.2 GLM-4.5
  61. 81.2 Kimi K2 Thinking
  62. 81.2 Qwen3.5-122B-A10B
  63. 81.2 GLM-5.2
  64. 80.7 Claude Sonnet 4
  65. 80.7 Kimi K2
  66. 80.7 Claude Opus 4.5
  67. 80.7 GPT-5.2
  68. 80.7 Claude Opus 4.6
  69. 80.7 Qwen3.5-Plus
  70. 80.7 Claude Sonnet 4.6
  71. 80.7 Qwen 3.6 Plus
  72. 80.7 Gemma 4 31B IT
  73. 80.7 Nemotron 3 Ultra
  74. 80.3 Gemini 2.5 Flash
  75. 80.3 Grok 4.1 Fast
  76. 80.3 GLM-4.7
  77. 80.3 GPT-5.6 Luna
  78. 80.3 DeepSeek V4.1 Flash 1017
  79. 79.8 Gemini 3.1 Flash-Lite
  80. 79.4 o1
  81. 79.4 GPT-4.1
  82. 79.4 GPT-5 mini
  83. 79.4 Mistral Large 3
  84. 79.4 MiMo-V2-Pro
  85. 79.4 MiniMax-M3
  86. 79.4 Muse Glimmer
  87. 79.4 Qwen 3.8 27B
  88. 79.4 Claude Haiku 5.5
  89. 78.9 Claude 3.7 Sonnet
  90. 78.9 Mistral Medium 3.1
  91. 78.9 GLM-4.6
  92. 78.9 Qwen3.5-27B
  93. 78.9 MiMo-V2.6 Pro
  94. 78.4 o4-mini
  95. 78.4 DeepSeek-R1
  96. 78.4 Kimi K2.5
  97. 78.4 MiMo-V2.6 Flash
  98. 78.0 GPT-4.5
  99. 78.0 Qwen3.5-35B-A3B
  100. 78.0 GPT-5.4 Mini
  101. 78.0 Qwen3.6 27B
  102. 78.0 Mistral Large 4
  103. 77.5 Grok 4 Fast
  104. 77.5 Qwen3.5-Flash
  105. 77.5 MiniMax-M2.7
  106. 77.5 Qwen 3.6 35B-A3B
  107. 77.5 MiMo-V2.5
  108. 77.5 Qwen3.7 Flash
  109. 77.1 Grok-2
  110. 77.1 Mistral Medium 3
  111. 77.1 Gemma 4 26B A4B
  112. 77.1 Gemma 4 12B
  113. 76.6 Qwen 3.6 Flash
  114. 76.6 Mistral Medium 3.5
  115. 76.6 Gemini 3.5 Flash-Lite
  116. 76.2 MiniMax-M2.1
  117. 75.7 Gemini 2.0 Flash
  118. 75.7 Grok-3 mini
  119. 75.7 Claude Haiku 4.5
  120. 75.7 MiMo-V2.5-Pro
  121. 75.2 Llama 4 Maverick
  122. 75.2 GPT-4.1 mini
  123. 75.2 Qwen3-235B-A22B
  124. 75.2 Qwen3-30B-A3B
  125. 75.2 Qwen3-32B
  126. 74.8 Magistral Medium 1.0
  127. 74.3 Nemotron 3.5 Lightning 30B-A3B
  128. 73.8 DeepSeek-V3
  129. 73.8 GPT-4o
  130. 73.8 gpt-oss-120b
  131. 73.4 Gemma 2 27B
  132. 73.4 Gemma 3 27B
  133. 73.4 Gemini 2.5 Flash-Lite
  134. 73.4 Command A+
  135. 72.9 Doubao-1.5-pro
  136. 72.9 LFM2.5-24B-A2B
  137. 72.5 Llama 3.1-405B
  138. 72.5 o3-mini
  139. 72.5 DeepSeek-V3.2
  140. 72.5 Qwen3.5-9B
  141. 72.5 GPT-5.4 Nano
  142. 72.0 Gemini 1.5 Pro
  143. 72.0 Cohere Command A
  144. 72.0 Grok 3
  145. 71.6 Gemini 2.0 Flash-Lite
  146. 71.6 MiniMax-M3.1-Flash
  147. 71.1 learnlm-1.5-pro-experimental
  148. 71.1 GPT-4o mini
  149. 71.1 Mistral Large 2
  150. 71.1 Amazon Nova Pro
  151. 70.6 Qwen2.5-72B
  152. 70.6 Llama 3.2 90B
  153. 70.6 GPT-5 nano
  154. 70.6 Gemma 4 E4B
  155. 70.2 GPT-4 Turbo
  156. 70.2 Llama 3.1-70B
  157. 70.2 Qwen3-14B
  158. 70.2 gpt-oss-20b
  159. 70.2 DeepSeek-V3.1
  160. 69.7 Llama 3.3 70B
  161. 69.7 Olmo 3.1 32B Think
  162. 69.3 Phi-4
  163. 69.3 Mistral Small 3.1
  164. 69.3 Qwen3-8B
  165. 69.3 Mistral Small 3.2
  166. 68.8 GPT-4
  167. 68.3 Claude 3 Opus
  168. 68.3 Gemini 1.5 Flash
  169. 68.3 o1-mini
  170. 68.3 Amazon Nova Lite
  171. 68.3 Mistral Small 3
  172. 68.3 glm-4.7-flash
  173. 67.9 Llama 4 Scout
  174. 67.9 Nova 2 Lite
  175. 67.4 Jamba 1.5 Large
  176. 67.4 Jamba 1.7 Large
  177. 67.4 Nemotron 3 Nano 30B-A3B
  178. 67.4 MiMo-V2-Flash
  179. 67.4 Granite 4.2 8B
  180. 67.0 Claude 3.5 Haiku
  181. 67.0 Ministral 3 14B
  182. 66.1 Claude 3.5 Sonnet
  183. 66.1 Command R+
  184. 66.1 Claude 3.5 Sonnet (October 2024)
  185. 66.1 Jamba 1.6 Large
  186. 66.1 Qwen3-4B
  187. 65.6 Yi-Large
  188. 65.6 Gemma 3 12B
  189. 65.6 Mistral Small 4
  190. 65.1 Llama 3.2 11B
  191. 65.1 sonar
  192. 64.7 Mixtral 8x7B
  193. 64.7 Llama 3.1-8B
  194. 64.2 GPT-4.1 nano
  195. 63.8 Yi-Lightning
  196. 63.8 Qwen3-1.7B
  197. 63.8 Olmo 3.1 32B Instruct
  198. 63.3 Phi-3.5-vision-instruct
  199. 62.4 Mixtral 8x22B
  200. 62.4 OLMo 2 32B Instruct
  201. 62.4 Gemma 3n E4B
  202. 61.9 Mistral NeMo
  203. 61.9 Qwen2.5-7B
  204. 61.0 Phi-3.5-mini
  205. 60.5 Amazon Nova Micro
  206. 60.5 Jamba 1.6 Mini
  207. 60.5 Gemma 3n E2B
  208. 59.6 Ministral 8B
  209. 59.6 Gemma 4 E2B
  210. 59.2 Ministral 3B
  211. 59.2 LFM2-8B-A1B
  212. 59.2 LFM3-1.7B
  213. 57.8 Phi-4-multimodal
  214. 57.3 Gemini 1.5 Flash 8B
  215. 56.9 Jamba 1.5 Mini
  216. 56.0 Claude 3 Haiku
  217. 55.0 Gemma 3 4B
  218. 54.1 Jamba 1.7 Mini
  219. 53.2 Command R7B
  220. 52.3 GPT-3.5 Turbo
  221. 52.3 LFM2-1.2B
  222. 47.3 Qwen3-0.6B
  223. 45.4 LFM2.5-1.2B-Thinking
  224. 38.1 Gemma 3 1B
  225. 28.4 Llama 3.2 1B
  226. 28.0 SmolLM2-1.7B
  227. 28.0 LFM2.5-1.2B-Instruct
  228. 26.6 OLMo 2 1B Instruct
  229. 26.1 SmolLM2-360M
  230. 24.8 SmolLM2-135M
  1. 1 Claude Fable 5.1 89.5%
  2. 2 Gemini 4 Argon high 89.5%
  3. 3 Claude Opus 5 89%
  4. 4 Gemini 3.6 Flash 88.5%
  5. 5 GPT-5.5 88.1%
  6. 6 Qwen 3.8 Max 88.1%
  7. 7 Muse Spark 1.2 88.1%
  8. 8 GPT-6 Astra high 88.1%
  9. 9 GPT-5.6 Sol 87.6%
  10. 10 GPT-6.1 Terra medium 87.6%
  11. 11 Claude Fable 5 87.2%
  12. 12 Gemini 3.7 Flash 87.2%
  13. 13 Qwen3.8 Max (0902) 87.2%
  14. 14 Gemini 3.1 Pro 86.7%
  15. 15 Muse Wonder high 86.7%
  16. 16 Gemini 3.5 Flash 86.2%
  17. 17 Grok 4.6 86.2%
  18. 18 Qwen3.8-2.4T-A95B 86.2%
  19. 19 Gemini 3.8 Flash 86.2%
  20. 20 GPT-6.1 Sol high 86.2%
  21. 21 Gemini 2.5 Pro Jun 2025 85.8%
  22. 22 Gemini 3 Flash 85.8%
  23. 23 Claude Opus 4.7 85.3%
  24. 24 Qwen3.7-Max 85.3%
  25. 25 Claude Opus 5.5 medium 85.3%
  26. 26 Grok 4.5 84.9%
  27. 27 Muse Spark 1.3 84.9%
  28. 28 Gemini 3 Pro 84.4%
  29. 29 Qwen3.8 Flash 84.4%
  30. 30 Grok 4.7 low 84.4%
  31. 31 Claude Sonnet 5.5 medium 84.4%
  32. 32 Grok 4.20 reasoning 83.9%
  33. 33 GPT-5.4 medium 83.9%
  34. 34 Kimi K3.1 Pro 83.9%
  35. 35 Claude Opus 4 Thinking (low) 83.5%
  36. 36 Claude Sonnet 4.5 thinking 83.5%
  37. 37 Qwen3.5 397B-A17B 83.5%
  38. 38 Muse Spark 1.1 83.5%
  39. 39 Kimi K2.6 83%
  40. 40 DeepSeek-V4-Flash 83%
  41. 41 Claude Opus 4.8 83%
  42. 42 Inkling 83%
  43. 43 Tencent Hy4 preview 83%
  44. 44 GLM-5 82.6%
  45. 45 Qwen 3.6 Max (Preview) 82.6%
  46. 46 Claude Sonnet 5 82.6%
  47. 47 o3 82.1%
  48. 48 Grok 4 82.1%
  49. 49 GPT-5.1 medium 82.1%
  50. 50 DeepSeek-V4-Pro 82.1%
  51. 51 Grok 4.3 82.1%
  52. 52 GPT-5.6 Terra 82.1%
  53. 53 Inkling-Small 82.1%
  54. 54 GLM-5.3 82.1%
  55. 55 GPT-5 medium 81.7%
  56. 56 GLM-5.1 81.7%
  57. 57 Kimi K3 81.7%
  58. 58 DeepSeek V4.1 Flash 81.7%
  59. 59 Qwen3-235B-A22B-Thinking Jul 2025 81.2%
  60. 60 GLM-4.5 81.2%
  61. 61 Kimi K2 Thinking 81.2%
  62. 62 Qwen3.5-122B-A10B 81.2%
  63. 63 GLM-5.2 81.2%
  64. 64 Claude Sonnet 4 Thinking (low) 80.7%
  65. 65 Kimi K2 Sep 2025 80.7%
  66. 66 Claude Opus 4.5 80.7%
  67. 67 GPT-5.2 medium 80.7%
  68. 68 Claude Opus 4.6 80.7%
  69. 69 Qwen3.5-Plus 80.7%
  70. 70 Claude Sonnet 4.6 Adaptive Thinking 80.7%
  71. 71 Qwen 3.6 Plus 80.7%
  72. 72 Gemma 4 31B IT 80.7%
  73. 73 Nemotron 3 Ultra 80.7%
  74. 74 Gemini 2.5 Flash Sep 2025 80.3%
  75. 75 Grok 4.1 Fast 80.3%
  76. 76 GLM-4.7 80.3%
  77. 77 GPT-5.6 Luna 80.3%
  78. 78 DeepSeek V4.1 Flash 1017 80.3%
  79. 79 Gemini 3.1 Flash-Lite 79.8%
  80. 80 o1 medium 79.4%
  81. 81 GPT-4.1 79.4%
  82. 82 GPT-5 mini medium 79.4%
  83. 83 Mistral Large 3 79.4%
  84. 84 MiMo-V2-Pro 79.4%
  85. 85 MiniMax-M3 79.4%
  86. 86 Muse Glimmer 79.4%
  87. 87 Qwen 3.8 27B 79.4%
  88. 88 Claude Haiku 5.5 medium 79.4%
  89. 89 Claude 3.7 Sonnet Thinking (medium) 78.9%
  90. 90 Mistral Medium 3.1 78.9%
  91. 91 GLM-4.6 78.9%
  92. 92 Qwen3.5-27B 78.9%
  93. 93 MiMo-V2.6 Pro 78.9%
  94. 94 o4-mini 78.4%
  95. 95 DeepSeek-R1 May 2025 78.4%
  96. 96 Kimi K2.5 78.4%
  97. 97 MiMo-V2.6 Flash 78.4%
  98. 98 GPT-4.5 78%
  99. 99 Qwen3.5-35B-A3B 78%
  100. 100 GPT-5.4 Mini medium 78%
  101. 101 Qwen3.6 27B 78%
  102. 102 Mistral Large 4 78%
  103. 103 Grok 4 Fast 77.5%
  104. 104 Qwen3.5-Flash 77.5%
  105. 105 MiniMax-M2.7 77.5%
  106. 106 Qwen 3.6 35B-A3B 77.5%
  107. 107 MiMo-V2.5 77.5%
  108. 108 Qwen3.7 Flash 77.5%
  109. 109 Grok-2 Dec 2024 77.1%
  110. 110 Mistral Medium 3 77.1%
  111. 111 Gemma 4 26B A4B 77.1%
  112. 112 Gemma 4 12B 77.1%
  113. 113 Qwen 3.6 Flash 76.6%
  114. 114 Mistral Medium 3.5 76.6%
  115. 115 Gemini 3.5 Flash-Lite 76.6%
  116. 116 MiniMax-M2.1 76.2%
  117. 117 Gemini 2.0 Flash Feb 2025 75.7%
  118. 118 Grok-3 mini 75.7%
  119. 119 Claude Haiku 4.5 75.7%
  120. 120 MiMo-V2.5-Pro 75.7%
  121. 121 Llama 4 Maverick 75.2%
  122. 122 GPT-4.1 mini 75.2%
  123. 123 Qwen3-235B-A22B 75.2%
  124. 124 Qwen3-30B-A3B 75.2%
  125. 125 Qwen3-32B 75.2%
  126. 126 Magistral Medium 1.0 74.8%
  127. 127 Nemotron 3.5 Lightning 30B-A3B 74.3%
  128. 128 DeepSeek-V3 Mar 2025 73.8%
  129. 129 GPT-4o Mar 2025 73.8%
  130. 130 gpt-oss-120b 73.8%
  131. 131 Gemma 2 27B 73.4%
  132. 132 Gemma 3 27B 73.4%
  133. 133 Gemini 2.5 Flash-Lite Sep 2025 73.4%
  134. 134 Command A+ 73.4%
  135. 135 Doubao-1.5-pro 72.9%
  136. 136 LFM2.5-24B-A2B 72.9%
  137. 137 Llama 3.1-405B 72.5%
  138. 138 o3-mini 72.5%
  139. 139 DeepSeek-V3.2 72.5%
  140. 140 Qwen3.5-9B 72.5%
  141. 141 GPT-5.4 Nano medium 72.5%
  142. 142 Gemini 1.5 Pro Sep 2024 72%
  143. 143 Cohere Command A 72%
  144. 144 Grok 3 72%
  145. 145 Gemini 2.0 Flash-Lite 71.6%
  146. 146 MiniMax-M3.1-Flash 71.6%
  147. 147 learnlm-1.5-pro-experimental 71.1%
  148. 148 GPT-4o mini 71.1%
  149. 149 Mistral Large 2 Nov 2024 71.1%
  150. 150 Amazon Nova Pro 71.1%
  151. 151 Qwen2.5-72B 70.6%
  152. 152 Llama 3.2 90B 70.6%
  153. 153 GPT-5 nano medium 70.6%
  154. 154 Gemma 4 E4B 70.6%
  155. 155 GPT-4 Turbo Apr 2024 70.2%
  156. 156 Llama 3.1-70B 70.2%
  157. 157 Qwen3-14B 70.2%
  158. 158 gpt-oss-20b 70.2%
  159. 159 DeepSeek-V3.1 70.2%
  160. 160 Llama 3.3 70B 69.7%
  161. 161 Olmo 3.1 32B Think 69.7%
  162. 162 Phi-4 69.3%
  163. 163 Mistral Small 3.1 69.3%
  164. 164 Qwen3-8B 69.3%
  165. 165 Mistral Small 3.2 69.3%
  166. 166 GPT-4 Jun 2023 68.8%
  167. 167 Claude 3 Opus 68.3%
  168. 168 Gemini 1.5 Flash Sep 2024 68.3%
  169. 169 o1-mini 68.3%
  170. 170 Amazon Nova Lite 68.3%
  171. 171 Mistral Small 3 68.3%
  172. 172 glm-4.7-flash 68.3%
  173. 173 Llama 4 Scout 67.9%
  174. 174 Nova 2 Lite 67.9%
  175. 175 Jamba 1.5 Large 67.4%
  176. 176 Jamba 1.7 Large 67.4%
  177. 177 Nemotron 3 Nano 30B-A3B 67.4%
  178. 178 MiMo-V2-Flash 67.4%
  179. 179 Granite 4.2 8B 67.4%
  180. 180 Claude 3.5 Haiku 67%
  181. 181 Ministral 3 14B 67%
  182. 182 Claude 3.5 Sonnet 66.1%
  183. 183 Command R+ 66.1%
  184. 184 Claude 3.5 Sonnet (October 2024) 66.1%
  185. 185 Jamba 1.6 Large 66.1%
  186. 186 Qwen3-4B 66.1%
  187. 187 Yi-Large 65.6%
  188. 188 Gemma 3 12B 65.6%
  189. 189 Mistral Small 4 65.6%
  190. 190 Llama 3.2 11B 65.1%
  191. 191 sonar 65.1%
  192. 192 Mixtral 8x7B 64.7%
  193. 193 Llama 3.1-8B 64.7%
  194. 194 GPT-4.1 nano 64.2%
  195. 195 Yi-Lightning 63.8%
  196. 196 Qwen3-1.7B 63.8%
  197. 197 Olmo 3.1 32B Instruct 63.8%
  198. 198 Phi-3.5-vision-instruct 63.3%
  199. 199 Mixtral 8x22B 62.4%
  200. 200 OLMo 2 32B Instruct 62.4%
  201. 201 Gemma 3n E4B 62.4%
  202. 202 Mistral NeMo 61.9%
  203. 203 Qwen2.5-7B 61.9%
  204. 204 Phi-3.5-mini 61%
  205. 205 Amazon Nova Micro 60.5%
  206. 206 Jamba 1.6 Mini 60.5%
  207. 207 Gemma 3n E2B 60.5%
  208. 208 Ministral 8B 59.6%
  209. 209 Gemma 4 E2B 59.6%
  210. 210 Ministral 3B 59.2%
  211. 211 LFM2-8B-A1B 59.2%
  212. 212 LFM3-1.7B 59.2%
  213. 213 Phi-4-multimodal 57.8%
  214. 214 Gemini 1.5 Flash 8B 57.3%
  215. 215 Jamba 1.5 Mini 56.9%
  216. 216 Claude 3 Haiku 56%
  217. 217 Gemma 3 4B 55%
  218. 218 Jamba 1.7 Mini 54.1%
  219. 219 Command R7B 53.2%
  220. 220 GPT-3.5 Turbo Jan 2024 52.3%
  221. 221 LFM2-1.2B 52.3%
  222. 222 Qwen3-0.6B 47.3%
  223. 223 LFM2.5-1.2B-Thinking 45.4%
  224. 224 Gemma 3 1B 38.1%
  225. 225 Llama 3.2 1B 28.4%
  226. 226 SmolLM2-1.7B 28%
  227. 227 LFM2.5-1.2B-Instruct 28%
  228. 228 OLMo 2 1B Instruct 26.6%
  229. 229 SmolLM2-360M 26.1%
  230. 230 SmolLM2-135M 24.8%

Source: Official SEND Pedagogy Benchmark leaderboard

#EntryScoreSettingDateSource
1 Claude Fable 5.1 89.5% Sep 2026 Official leaderboard
2 Gemini 4 Argon 89.5% high Sep 2026 Official leaderboard
3 Claude Opus 5 89% Jul 2026 Official leaderboard
4 Gemini 3.6 Flash 88.5% Jul 2026 Official leaderboard
5 GPT-5.5 88.1% Apr 2026 Official leaderboard
6 Qwen 3.8 Max 88.1% Aug 2026 Official leaderboard
7 Muse Spark 1.2 88.1% Aug 2026 Official leaderboard
8 GPT-6 Astra 88.1% high Sep 2026 Official leaderboard
9 GPT-5.6 Sol 87.6% Jul 2026 Official leaderboard
10 GPT-6.1 Terra 87.6% medium Oct 2026 Official leaderboard
11 Claude Fable 5 87.2% Jun 2026 Official leaderboard
12 Gemini 3.7 Flash 87.2% Aug 2026 Official leaderboard
13 Qwen3.8 Max (0902) 87.2% Sep 2026 Official leaderboard
14 Gemini 3.1 Pro 86.7% Feb 2026 Official leaderboard
15 Muse Wonder 86.7% high Oct 2026 Official leaderboard
16 Gemini 3.5 Flash 86.2% May 2026 Official leaderboard
17 Grok 4.6 86.2% Aug 2026 Official leaderboard
18 Qwen3.8-2.4T-A95B 86.2% Aug 2026 Official leaderboard
19 Gemini 3.8 Flash 86.2% Sep 2026 Official leaderboard
20 GPT-6.1 Sol 86.2% high Sep 2026 Official leaderboard
21 Gemini 2.5 Pro Jun 2025 85.8% Jun 2025 Official leaderboard
22 Gemini 3 Flash 85.8% Dec 2025 Official leaderboard
23 Claude Opus 4.7 85.3% Apr 2026 Official leaderboard
24 Qwen3.7-Max 85.3% May 2026 Official leaderboard
25 Claude Opus 5.5 85.3% medium Sep 2026 Official leaderboard
26 Grok 4.5 84.9% Jul 2026 Official leaderboard
27 Muse Spark 1.3 84.9% Sep 2026 Official leaderboard
28 Gemini 3 Pro 84.4% Nov 2025 Official leaderboard
29 Qwen3.8 Flash 84.4% Aug 2026 Official leaderboard
30 Grok 4.7 84.4% low Sep 2026 Official leaderboard
31 Claude Sonnet 5.5 84.4% medium Sep 2026 Official leaderboard
32 Grok 4.20 83.9% reasoning Feb 2026 Official leaderboard
33 GPT-5.4 83.9% medium Mar 2026 Official leaderboard
34 Kimi K3.1 Pro 83.9% Oct 2026 Official leaderboard
35 Claude Opus 4 83.5% Thinking (low) May 2025 Official leaderboard
36 Claude Sonnet 4.5 83.5% thinking Sep 2025 Official leaderboard
37 Qwen3.5 397B-A17B 83.5% Feb 2026 Official leaderboard
38 Muse Spark 1.1 83.5% Jul 2026 Official leaderboard
39 Kimi K2.6 83% Apr 2026 Official leaderboard
40 DeepSeek-V4-Flash 83% Apr 2026 Official leaderboard
41 Claude Opus 4.8 83% May 2026 Official leaderboard
42 Inkling 83% Jul 2026 Official leaderboard
43 Tencent Hy4 preview 83% Aug 2026 Official leaderboard
44 GLM-5 82.6% Feb 2026 Official leaderboard
45 Qwen 3.6 Max (Preview) 82.6% Apr 2026 Official leaderboard
46 Claude Sonnet 5 82.6% Jun 2026 Official leaderboard
47 o3 82.1% Apr 2025 Official leaderboard
48 Grok 4 82.1% Jul 2025 Official leaderboard
49 GPT-5.1 82.1% medium Nov 2025 Official leaderboard
50 DeepSeek-V4-Pro 82.1% Apr 2026 Official leaderboard
51 Grok 4.3 82.1% Apr 2026 Official leaderboard
52 GPT-5.6 Terra 82.1% Jul 2026 Official leaderboard
53 Inkling-Small 82.1% Jul 2026 Official leaderboard
54 GLM-5.3 82.1% Aug 2026 Official leaderboard
55 GPT-5 81.7% medium Aug 2025 Official leaderboard
56 GLM-5.1 81.7% Apr 2026 Official leaderboard
57 Kimi K3 81.7% Jul 2026 Official leaderboard
58 DeepSeek V4.1 Flash 81.7% Sep 2026 Official leaderboard
59 Qwen3-235B-A22B-Thinking Jul 2025 81.2% Jul 2025 Official leaderboard
60 GLM-4.5 81.2% Aug 2025 Official leaderboard
61 Kimi K2 Thinking 81.2% Nov 2025 Official leaderboard
62 Qwen3.5-122B-A10B 81.2% Feb 2026 Official leaderboard
63 GLM-5.2 81.2% Jun 2026 Official leaderboard
64 Claude Sonnet 4 80.7% Thinking (low) May 2025 Official leaderboard
65 Kimi K2 Sep 2025 80.7% Sep 2025 Official leaderboard
66 Claude Opus 4.5 80.7% Nov 2025 Official leaderboard
67 GPT-5.2 80.7% medium Dec 2025 Official leaderboard
68 Claude Opus 4.6 80.7% Feb 2026 Official leaderboard
69 Qwen3.5-Plus 80.7% Feb 2026 Official leaderboard
70 Claude Sonnet 4.6 80.7% Adaptive Thinking Feb 2026 Official leaderboard
71 Qwen 3.6 Plus 80.7% Mar 2026 Official leaderboard
72 Gemma 4 31B IT 80.7% Apr 2026 Official leaderboard
73 Nemotron 3 Ultra 80.7% Jun 2026 Official leaderboard
74 Gemini 2.5 Flash Sep 2025 80.3% Sep 2025 Official leaderboard
75 Grok 4.1 Fast 80.3% Nov 2025 Official leaderboard
76 GLM-4.7 80.3% Dec 2025 Official leaderboard
77 GPT-5.6 Luna 80.3% Jul 2026 Official leaderboard
78 DeepSeek V4.1 Flash 1017 80.3% Oct 2026 Official leaderboard
79 Gemini 3.1 Flash-Lite 79.8% Mar 2026 Official leaderboard
80 o1 79.4% medium Dec 2024 Official leaderboard
81 GPT-4.1 79.4% Apr 2025 Official leaderboard
82 GPT-5 mini 79.4% medium Aug 2025 Official leaderboard
83 Mistral Large 3 79.4% Dec 2025 Official leaderboard
84 MiMo-V2-Pro 79.4% Mar 2026 Official leaderboard
85 MiniMax-M3 79.4% Jun 2026 Official leaderboard
86 Muse Glimmer 79.4% Aug 2026 Official leaderboard
87 Qwen 3.8 27B 79.4% Aug 2026 Official leaderboard
88 Claude Haiku 5.5 79.4% medium Oct 2026 Official leaderboard
89 Claude 3.7 Sonnet 78.9% Thinking (medium) Feb 2025 Official leaderboard
90 Mistral Medium 3.1 78.9% Aug 2025 Official leaderboard
91 GLM-4.6 78.9% Sep 2025 Official leaderboard
92 Qwen3.5-27B 78.9% Feb 2026 Official leaderboard
93 MiMo-V2.6 Pro 78.9% Sep 2026 Official leaderboard
94 o4-mini 78.4% Apr 2025 Official leaderboard
95 DeepSeek-R1 May 2025 78.4% May 2025 Official leaderboard
96 Kimi K2.5 78.4% Jan 2026 Official leaderboard
97 MiMo-V2.6 Flash 78.4% Sep 2026 Official leaderboard
98 GPT-4.5 78% Feb 2025 Official leaderboard
99 Qwen3.5-35B-A3B 78% Feb 2026 Official leaderboard
100 GPT-5.4 Mini 78% medium Mar 2026 Official leaderboard
101 Qwen3.6 27B 78% Apr 2026 Official leaderboard
102 Mistral Large 4 78% Oct 2026 Official leaderboard
103 Grok 4 Fast 77.5% Sep 2025 Official leaderboard
104 Qwen3.5-Flash 77.5% Feb 2026 Official leaderboard
105 MiniMax-M2.7 77.5% Mar 2026 Official leaderboard
106 Qwen 3.6 35B-A3B 77.5% Apr 2026 Official leaderboard
107 MiMo-V2.5 77.5% Apr 2026 Official leaderboard
108 Qwen3.7 Flash 77.5% Jul 2026 Official leaderboard
109 Grok-2 Dec 2024 77.1% Dec 2024 Official leaderboard
110 Mistral Medium 3 77.1% May 2025 Official leaderboard
111 Gemma 4 26B A4B 77.1% Apr 2026 Official leaderboard
112 Gemma 4 12B 77.1% Jun 2026 Official leaderboard
113 Qwen 3.6 Flash 76.6% Apr 2026 Official leaderboard
114 Mistral Medium 3.5 76.6% Apr 2026 Official leaderboard
115 Gemini 3.5 Flash-Lite 76.6% Jul 2026 Official leaderboard
116 MiniMax-M2.1 76.2% Dec 2025 Official leaderboard
117 Gemini 2.0 Flash Feb 2025 75.7% Feb 2025 Official leaderboard
118 Grok-3 mini 75.7% Apr 2025 Official leaderboard
119 Claude Haiku 4.5 75.7% Oct 2025 Official leaderboard
120 MiMo-V2.5-Pro 75.7% Apr 2026 Official leaderboard
121 Llama 4 Maverick 75.2% Apr 2025 Official leaderboard
122 GPT-4.1 mini 75.2% Apr 2025 Official leaderboard
123 Qwen3-235B-A22B 75.2% Apr 2025 Official leaderboard
124 Qwen3-30B-A3B 75.2% Apr 2025 Official leaderboard
125 Qwen3-32B 75.2% Apr 2025 Official leaderboard
126 Magistral Medium 1.0 74.8% Jun 2025 Official leaderboard
127 Nemotron 3.5 Lightning 30B-A3B 74.3% Aug 2026 Official leaderboard
128 DeepSeek-V3 Mar 2025 73.8% Mar 2025 Official leaderboard
129 GPT-4o Mar 2025 73.8% Mar 2025 Official leaderboard
130 gpt-oss-120b 73.8% Aug 2025 Official leaderboard
131 Gemma 2 27B 73.4% Jun 2024 Official leaderboard
132 Gemma 3 27B 73.4% Mar 2025 Official leaderboard
133 Gemini 2.5 Flash-Lite Sep 2025 73.4% Sep 2025 Official leaderboard
134 Command A+ 73.4% May 2026 Official leaderboard
135 Doubao-1.5-pro 72.9% Jan 2025 Official leaderboard
136 LFM2.5-24B-A2B 72.9% Oct 2026 Official leaderboard
137 Llama 3.1-405B 72.5% Jul 2024 Official leaderboard
138 o3-mini 72.5% Jan 2025 Official leaderboard
139 DeepSeek-V3.2 72.5% Dec 2025 Official leaderboard
140 Qwen3.5-9B 72.5% Feb 2026 Official leaderboard
141 GPT-5.4 Nano 72.5% medium Mar 2026 Official leaderboard
142 Gemini 1.5 Pro Sep 2024 72% Sep 2024 Official leaderboard
143 Cohere Command A 72% Mar 2025 Official leaderboard
144 Grok 3 72% Apr 2025 Official leaderboard
145 Gemini 2.0 Flash-Lite 71.6% Feb 2025 Official leaderboard
146 MiniMax-M3.1-Flash 71.6% Oct 2026 Official leaderboard
147 learnlm-1.5-pro-experimental 71.1% Official leaderboard
148 Ministral 3 8B 71.1% Official leaderboard
149 GPT-4o mini 71.1% Jul 2024 Official leaderboard
150 Mistral Large 2 Nov 2024 71.1% Nov 2024 Official leaderboard
151 Amazon Nova Pro 71.1% Dec 2024 Official leaderboard
152 Qwen2.5-72B 70.6% Sep 2024 Official leaderboard
153 Llama 3.2 90B 70.6% Sep 2024 Official leaderboard
154 GPT-5 nano 70.6% medium Aug 2025 Official leaderboard
155 Gemma 4 E4B 70.6% Mar 2026 Official leaderboard
156 GPT-4 Turbo Apr 2024 70.2% Apr 2024 Official leaderboard
157 Llama 3.1-70B 70.2% Jul 2024 Official leaderboard
158 Qwen3-14B 70.2% Apr 2025 Official leaderboard
159 gpt-oss-20b 70.2% Aug 2025 Official leaderboard
160 DeepSeek-V3.1 70.2% Aug 2025 Official leaderboard
161 Hunyuan Large 69.7% Official leaderboard
162 Llama 3.3 70B 69.7% Dec 2024 Official leaderboard
163 Olmo 3.1 32B Think 69.7% Dec 2025 Official leaderboard
164 Phi-4 69.3% Dec 2024 Official leaderboard
165 Mistral Small 3.1 69.3% Mar 2025 Official leaderboard
166 Qwen3-8B 69.3% Apr 2025 Official leaderboard
167 Mistral Small 3.2 69.3% Jun 2025 Official leaderboard
168 GPT-4 Jun 2023 68.8% Jun 2023 Official leaderboard
169 Doubao-1.5 Lite 68.3% Official leaderboard
170 Claude 3 Opus 68.3% Feb 2024 Official leaderboard
171 Gemini 1.5 Flash Sep 2024 68.3% Aug 2024 Official leaderboard
172 o1-mini 68.3% Sep 2024 Official leaderboard
173 Amazon Nova Lite 68.3% Dec 2024 Official leaderboard
174 Mistral Small 3 68.3% Jan 2025 Official leaderboard
175 glm-4.7-flash 68.3% Jan 2026 Official leaderboard
176 Llama 4 Scout 67.9% Apr 2025 Official leaderboard
177 Nova 2 Lite 67.9% Dec 2025 Official leaderboard
178 Hunyuan Large (Long Context) 67.4% Official leaderboard
179 Jamba 1.5 Large 67.4% Aug 2024 Official leaderboard
180 Jamba 1.7 Large 67.4% Jul 2025 Official leaderboard
181 Nemotron 3 Nano 30B-A3B 67.4% Dec 2025 Official leaderboard
182 MiMo-V2-Flash 67.4% Dec 2025 Official leaderboard
183 Granite 4.2 8B 67.4% Aug 2026 Official leaderboard
184 Grok Beta 67% Official leaderboard
185 Claude 3.5 Haiku 67% Oct 2024 Official leaderboard
186 Ministral 3 14B 67% Dec 2025 Official leaderboard
187 Claude 3.5 Sonnet 66.1% Jun 2024 Official leaderboard
188 Command R+ 66.1% Aug 2024 Official leaderboard
189 Claude 3.5 Sonnet (October 2024) 66.1% Oct 2024 Official leaderboard
190 Jamba 1.6 Large 66.1% Mar 2025 Official leaderboard
191 Qwen3-4B 66.1% Apr 2025 Official leaderboard
192 Yi-Large 65.6% May 2024 Official leaderboard
193 Gemma 3 12B 65.6% Mar 2025 Official leaderboard
194 Mistral Small 4 65.6% Mar 2026 Official leaderboard
195 Llama 3.2 11B 65.1% Sep 2024 Official leaderboard
196 sonar 65.1% Jan 2025 Official leaderboard
197 Mixtral 8x7B 64.7% Dec 2023 Official leaderboard
198 Llama 3.1-8B 64.7% Jul 2024 Official leaderboard
199 GPT-4.1 nano 64.2% Apr 2025 Official leaderboard
200 Yi-Lightning 63.8% Dec 2024 Official leaderboard
201 Qwen3-1.7B 63.8% Apr 2025 Official leaderboard
202 Olmo 3.1 32B Instruct 63.8% Dec 2025 Official leaderboard
203 Olmo-3 7B Think 63.3% Official leaderboard
204 Phi-3.5-vision-instruct 63.3% Aug 2024 Official leaderboard
205 Mixtral 8x22B 62.4% Apr 2024 Official leaderboard
206 OLMo 2 32B Instruct 62.4% Mar 2025 Official leaderboard
207 Gemma 3n E4B 62.4% Jun 2025 Official leaderboard
208 Mistral NeMo 61.9% Jul 2024 Official leaderboard
209 Qwen2.5-7B 61.9% Sep 2024 Official leaderboard
210 Phi-3.5-mini 61% Aug 2024 Official leaderboard
211 SmolLM3-3B 60.5% Official leaderboard
212 Amazon Nova Micro 60.5% Dec 2024 Official leaderboard
213 Jamba 1.6 Mini 60.5% Mar 2025 Official leaderboard
214 Gemma 3n E2B 60.5% Jun 2025 Official leaderboard
215 LFM-3B 60.1% Official leaderboard
216 Ministral 3 3B 59.6% Official leaderboard
217 Ministral 8B 59.6% Oct 2024 Official leaderboard
218 Gemma 4 E2B 59.6% Mar 2026 Official leaderboard
219 Ministral 3B 59.2% Oct 2024 Official leaderboard
220 LFM2-8B-A1B 59.2% Oct 2025 Official leaderboard
221 LFM3-1.7B 59.2% Oct 2026 Official leaderboard
222 LFM-7B 58.3% Official leaderboard
223 Jamba Instruct 57.8% Official leaderboard
224 LFM2-2.6B 57.8% Official leaderboard
225 Phi-4-multimodal 57.8% Feb 2025 Official leaderboard
226 Gemini 1.5 Flash 8B 57.3% Oct 2024 Official leaderboard
227 Olmo-2 13B 56.9% Official leaderboard
228 Jamba 1.5 Mini 56.9% Aug 2024 Official leaderboard
229 Claude 3 Haiku 56% Mar 2024 Official leaderboard
230 Gemma 3 4B 55% Mar 2025 Official leaderboard
231 Jamba 1.7 Mini 54.1% Jul 2025 Official leaderboard
232 Command R7B 53.2% Dec 2024 Official leaderboard
233 GPT-3.5 Turbo Jan 2024 52.3% Jan 2024 Official leaderboard
234 LFM2-1.2B 52.3% Jul 2025 Official leaderboard
235 Olmo-3 7B 51.4% Official leaderboard
236 Qwen3-0.6B 47.3% Apr 2025 Official leaderboard
237 Olmo-2 7B 46.3% Official leaderboard
238 LFM2.5-1.2B-Thinking 45.4% Jan 2026 Official leaderboard
239 Gemma 3 1B 38.1% Mar 2025 Official leaderboard
240 LFM2-700M 33.9% Official leaderboard
241 LFM2-350M 31.6% Official leaderboard
242 Llama 3.2 1B 28.4% Sep 2024 Official leaderboard
243 SmolLM2-1.7B 28% Oct 2024 Official leaderboard
244 LFM2.5-1.2B-Instruct 28% Jan 2026 Official leaderboard
245 OLMo 2 1B Instruct 26.6% Apr 2025 Official leaderboard
246 SmolLM2-360M 26.1% Oct 2024 Official leaderboard
247 SmolLM2-135M 24.8% Oct 2024 Official leaderboard
#ModelScoreSettingDateSource
1 Claude Fable 5.1 89.5% Sep 2026 Official leaderboard
2 Gemini 4 Argon 89.5% high Sep 2026 Official leaderboard
3 Claude Opus 5 89% Jul 2026 Official leaderboard
4 Gemini 3.6 Flash 88.5% Jul 2026 Official leaderboard
5 GPT-5.5 88.1% Apr 2026 Official leaderboard
6 Qwen 3.8 Max 88.1% Aug 2026 Official leaderboard
7 Muse Spark 1.2 88.1% Aug 2026 Official leaderboard
8 GPT-6 Astra 88.1% high Sep 2026 Official leaderboard
9 GPT-5.6 Sol 87.6% Jul 2026 Official leaderboard
10 GPT-6.1 Terra 87.6% medium Oct 2026 Official leaderboard
11 Claude Fable 5 87.2% Jun 2026 Official leaderboard
12 Gemini 3.7 Flash 87.2% Aug 2026 Official leaderboard
13 Qwen3.8 Max (0902) 87.2% Sep 2026 Official leaderboard
14 Gemini 3.1 Pro 86.7% Feb 2026 Official leaderboard
15 Muse Wonder 86.7% high Oct 2026 Official leaderboard
16 Gemini 3.5 Flash 86.2% May 2026 Official leaderboard
17 Grok 4.6 86.2% Aug 2026 Official leaderboard
18 Qwen3.8-2.4T-A95B 86.2% Aug 2026 Official leaderboard
19 Gemini 3.8 Flash 86.2% Sep 2026 Official leaderboard
20 GPT-6.1 Sol 86.2% high Sep 2026 Official leaderboard
21 Gemini 2.5 Pro Jun 2025 85.8% Jun 2025 Official leaderboard
22 Gemini 3 Flash 85.8% Dec 2025 Official leaderboard
23 Claude Opus 4.7 85.3% Apr 2026 Official leaderboard
24 Qwen3.7-Max 85.3% May 2026 Official leaderboard
25 Claude Opus 5.5 85.3% medium Sep 2026 Official leaderboard
26 Grok 4.5 84.9% Jul 2026 Official leaderboard
27 Muse Spark 1.3 84.9% Sep 2026 Official leaderboard
28 Gemini 3 Pro 84.4% Nov 2025 Official leaderboard
29 Qwen3.8 Flash 84.4% Aug 2026 Official leaderboard
30 Grok 4.7 84.4% low Sep 2026 Official leaderboard
31 Claude Sonnet 5.5 84.4% medium Sep 2026 Official leaderboard
32 Grok 4.20 83.9% reasoning Feb 2026 Official leaderboard
33 GPT-5.4 83.9% medium Mar 2026 Official leaderboard
34 Kimi K3.1 Pro 83.9% Oct 2026 Official leaderboard
35 Claude Opus 4 83.5% Thinking (low) May 2025 Official leaderboard
36 Claude Sonnet 4.5 83.5% thinking Sep 2025 Official leaderboard
37 Qwen3.5 397B-A17B 83.5% Feb 2026 Official leaderboard
38 Muse Spark 1.1 83.5% Jul 2026 Official leaderboard
39 Kimi K2.6 83% Apr 2026 Official leaderboard
40 DeepSeek-V4-Flash 83% Apr 2026 Official leaderboard
41 Claude Opus 4.8 83% May 2026 Official leaderboard
42 Inkling 83% Jul 2026 Official leaderboard
43 Tencent Hy4 preview 83% Aug 2026 Official leaderboard
44 GLM-5 82.6% Feb 2026 Official leaderboard
45 Qwen 3.6 Max (Preview) 82.6% Apr 2026 Official leaderboard
46 Claude Sonnet 5 82.6% Jun 2026 Official leaderboard
47 o3 82.1% Apr 2025 Official leaderboard
48 Grok 4 82.1% Jul 2025 Official leaderboard
49 GPT-5.1 82.1% medium Nov 2025 Official leaderboard
50 DeepSeek-V4-Pro 82.1% Apr 2026 Official leaderboard
51 Grok 4.3 82.1% Apr 2026 Official leaderboard
52 GPT-5.6 Terra 82.1% Jul 2026 Official leaderboard
53 Inkling-Small 82.1% Jul 2026 Official leaderboard
54 GLM-5.3 82.1% Aug 2026 Official leaderboard
55 GPT-5 81.7% medium Aug 2025 Official leaderboard
56 GLM-5.1 81.7% Apr 2026 Official leaderboard
57 Kimi K3 81.7% Jul 2026 Official leaderboard
58 DeepSeek V4.1 Flash 81.7% Sep 2026 Official leaderboard
59 Qwen3-235B-A22B-Thinking Jul 2025 81.2% Jul 2025 Official leaderboard
60 GLM-4.5 81.2% Aug 2025 Official leaderboard
61 Kimi K2 Thinking 81.2% Nov 2025 Official leaderboard
62 Qwen3.5-122B-A10B 81.2% Feb 2026 Official leaderboard
63 GLM-5.2 81.2% Jun 2026 Official leaderboard
64 Claude Sonnet 4 80.7% Thinking (low) May 2025 Official leaderboard
65 Kimi K2 Sep 2025 80.7% Sep 2025 Official leaderboard
66 Claude Opus 4.5 80.7% Nov 2025 Official leaderboard
67 GPT-5.2 80.7% medium Dec 2025 Official leaderboard
68 Claude Opus 4.6 80.7% Feb 2026 Official leaderboard
69 Qwen3.5-Plus 80.7% Feb 2026 Official leaderboard
70 Claude Sonnet 4.6 80.7% Adaptive Thinking Feb 2026 Official leaderboard
71 Qwen 3.6 Plus 80.7% Mar 2026 Official leaderboard
72 Gemma 4 31B IT 80.7% Apr 2026 Official leaderboard
73 Nemotron 3 Ultra 80.7% Jun 2026 Official leaderboard
74 Gemini 2.5 Flash Sep 2025 80.3% Sep 2025 Official leaderboard
75 Grok 4.1 Fast 80.3% Nov 2025 Official leaderboard
76 GLM-4.7 80.3% Dec 2025 Official leaderboard
77 GPT-5.6 Luna 80.3% Jul 2026 Official leaderboard
78 DeepSeek V4.1 Flash 1017 80.3% Oct 2026 Official leaderboard
79 Gemini 3.1 Flash-Lite 79.8% Mar 2026 Official leaderboard
80 o1 79.4% medium Dec 2024 Official leaderboard
81 GPT-4.1 79.4% Apr 2025 Official leaderboard
82 GPT-5 mini 79.4% medium Aug 2025 Official leaderboard
83 Mistral Large 3 79.4% Dec 2025 Official leaderboard
84 MiMo-V2-Pro 79.4% Mar 2026 Official leaderboard
85 MiniMax-M3 79.4% Jun 2026 Official leaderboard
86 Muse Glimmer 79.4% Aug 2026 Official leaderboard
87 Qwen 3.8 27B 79.4% Aug 2026 Official leaderboard
88 Claude Haiku 5.5 79.4% medium Oct 2026 Official leaderboard
89 Claude 3.7 Sonnet 78.9% Thinking (medium) Feb 2025 Official leaderboard
90 Mistral Medium 3.1 78.9% Aug 2025 Official leaderboard
91 GLM-4.6 78.9% Sep 2025 Official leaderboard
92 Qwen3.5-27B 78.9% Feb 2026 Official leaderboard
93 MiMo-V2.6 Pro 78.9% Sep 2026 Official leaderboard
94 o4-mini 78.4% Apr 2025 Official leaderboard
95 DeepSeek-R1 May 2025 78.4% May 2025 Official leaderboard
96 Kimi K2.5 78.4% Jan 2026 Official leaderboard
97 MiMo-V2.6 Flash 78.4% Sep 2026 Official leaderboard
98 GPT-4.5 78% Feb 2025 Official leaderboard
99 Qwen3.5-35B-A3B 78% Feb 2026 Official leaderboard
100 GPT-5.4 Mini 78% medium Mar 2026 Official leaderboard
101 Qwen3.6 27B 78% Apr 2026 Official leaderboard
102 Mistral Large 4 78% Oct 2026 Official leaderboard
103 Grok 4 Fast 77.5% Sep 2025 Official leaderboard
104 Qwen3.5-Flash 77.5% Feb 2026 Official leaderboard
105 MiniMax-M2.7 77.5% Mar 2026 Official leaderboard
106 Qwen 3.6 35B-A3B 77.5% Apr 2026 Official leaderboard
107 MiMo-V2.5 77.5% Apr 2026 Official leaderboard
108 Qwen3.7 Flash 77.5% Jul 2026 Official leaderboard
109 Grok-2 Dec 2024 77.1% Dec 2024 Official leaderboard
110 Mistral Medium 3 77.1% May 2025 Official leaderboard
111 Gemma 4 26B A4B 77.1% Apr 2026 Official leaderboard
112 Gemma 4 12B 77.1% Jun 2026 Official leaderboard
113 Qwen 3.6 Flash 76.6% Apr 2026 Official leaderboard
114 Mistral Medium 3.5 76.6% Apr 2026 Official leaderboard
115 Gemini 3.5 Flash-Lite 76.6% Jul 2026 Official leaderboard
116 MiniMax-M2.1 76.2% Dec 2025 Official leaderboard
117 Gemini 2.0 Flash Feb 2025 75.7% Feb 2025 Official leaderboard
118 Grok-3 mini 75.7% Apr 2025 Official leaderboard
119 Claude Haiku 4.5 75.7% Oct 2025 Official leaderboard
120 MiMo-V2.5-Pro 75.7% Apr 2026 Official leaderboard
121 Llama 4 Maverick 75.2% Apr 2025 Official leaderboard
122 GPT-4.1 mini 75.2% Apr 2025 Official leaderboard
123 Qwen3-235B-A22B 75.2% Apr 2025 Official leaderboard
124 Qwen3-30B-A3B 75.2% Apr 2025 Official leaderboard
125 Qwen3-32B 75.2% Apr 2025 Official leaderboard
126 Magistral Medium 1.0 74.8% Jun 2025 Official leaderboard
127 Nemotron 3.5 Lightning 30B-A3B 74.3% Aug 2026 Official leaderboard
128 DeepSeek-V3 Mar 2025 73.8% Mar 2025 Official leaderboard
129 GPT-4o Mar 2025 73.8% Mar 2025 Official leaderboard
130 gpt-oss-120b 73.8% Aug 2025 Official leaderboard
131 Gemma 2 27B 73.4% Jun 2024 Official leaderboard
132 Gemma 3 27B 73.4% Mar 2025 Official leaderboard
133 Gemini 2.5 Flash-Lite Sep 2025 73.4% Sep 2025 Official leaderboard
134 Command A+ 73.4% May 2026 Official leaderboard
135 Doubao-1.5-pro 72.9% Jan 2025 Official leaderboard
136 LFM2.5-24B-A2B 72.9% Oct 2026 Official leaderboard
137 Llama 3.1-405B 72.5% Jul 2024 Official leaderboard
138 o3-mini 72.5% Jan 2025 Official leaderboard
139 DeepSeek-V3.2 72.5% Dec 2025 Official leaderboard
140 Qwen3.5-9B 72.5% Feb 2026 Official leaderboard
141 GPT-5.4 Nano 72.5% medium Mar 2026 Official leaderboard
142 Gemini 1.5 Pro Sep 2024 72% Sep 2024 Official leaderboard
143 Cohere Command A 72% Mar 2025 Official leaderboard
144 Grok 3 72% Apr 2025 Official leaderboard
145 Gemini 2.0 Flash-Lite 71.6% Feb 2025 Official leaderboard
146 MiniMax-M3.1-Flash 71.6% Oct 2026 Official leaderboard
147 learnlm-1.5-pro-experimental 71.1% Official leaderboard
148 GPT-4o mini 71.1% Jul 2024 Official leaderboard
149 Mistral Large 2 Nov 2024 71.1% Nov 2024 Official leaderboard
150 Amazon Nova Pro 71.1% Dec 2024 Official leaderboard
151 Qwen2.5-72B 70.6% Sep 2024 Official leaderboard
152 Llama 3.2 90B 70.6% Sep 2024 Official leaderboard
153 GPT-5 nano 70.6% medium Aug 2025 Official leaderboard
154 Gemma 4 E4B 70.6% Mar 2026 Official leaderboard
155 GPT-4 Turbo Apr 2024 70.2% Apr 2024 Official leaderboard
156 Llama 3.1-70B 70.2% Jul 2024 Official leaderboard
157 Qwen3-14B 70.2% Apr 2025 Official leaderboard
158 gpt-oss-20b 70.2% Aug 2025 Official leaderboard
159 DeepSeek-V3.1 70.2% Aug 2025 Official leaderboard
160 Llama 3.3 70B 69.7% Dec 2024 Official leaderboard
161 Olmo 3.1 32B Think 69.7% Dec 2025 Official leaderboard
162 Phi-4 69.3% Dec 2024 Official leaderboard
163 Mistral Small 3.1 69.3% Mar 2025 Official leaderboard
164 Qwen3-8B 69.3% Apr 2025 Official leaderboard
165 Mistral Small 3.2 69.3% Jun 2025 Official leaderboard
166 GPT-4 Jun 2023 68.8% Jun 2023 Official leaderboard
167 Claude 3 Opus 68.3% Feb 2024 Official leaderboard
168 Gemini 1.5 Flash Sep 2024 68.3% Aug 2024 Official leaderboard
169 o1-mini 68.3% Sep 2024 Official leaderboard
170 Amazon Nova Lite 68.3% Dec 2024 Official leaderboard
171 Mistral Small 3 68.3% Jan 2025 Official leaderboard
172 glm-4.7-flash 68.3% Jan 2026 Official leaderboard
173 Llama 4 Scout 67.9% Apr 2025 Official leaderboard
174 Nova 2 Lite 67.9% Dec 2025 Official leaderboard
175 Jamba 1.5 Large 67.4% Aug 2024 Official leaderboard
176 Jamba 1.7 Large 67.4% Jul 2025 Official leaderboard
177 Nemotron 3 Nano 30B-A3B 67.4% Dec 2025 Official leaderboard
178 MiMo-V2-Flash 67.4% Dec 2025 Official leaderboard
179 Granite 4.2 8B 67.4% Aug 2026 Official leaderboard
180 Claude 3.5 Haiku 67% Oct 2024 Official leaderboard
181 Ministral 3 14B 67% Dec 2025 Official leaderboard
182 Claude 3.5 Sonnet 66.1% Jun 2024 Official leaderboard
183 Command R+ 66.1% Aug 2024 Official leaderboard
184 Claude 3.5 Sonnet (October 2024) 66.1% Oct 2024 Official leaderboard
185 Jamba 1.6 Large 66.1% Mar 2025 Official leaderboard
186 Qwen3-4B 66.1% Apr 2025 Official leaderboard
187 Yi-Large 65.6% May 2024 Official leaderboard
188 Gemma 3 12B 65.6% Mar 2025 Official leaderboard
189 Mistral Small 4 65.6% Mar 2026 Official leaderboard
190 Llama 3.2 11B 65.1% Sep 2024 Official leaderboard
191 sonar 65.1% Jan 2025 Official leaderboard
192 Mixtral 8x7B 64.7% Dec 2023 Official leaderboard
193 Llama 3.1-8B 64.7% Jul 2024 Official leaderboard
194 GPT-4.1 nano 64.2% Apr 2025 Official leaderboard
195 Yi-Lightning 63.8% Dec 2024 Official leaderboard
196 Qwen3-1.7B 63.8% Apr 2025 Official leaderboard
197 Olmo 3.1 32B Instruct 63.8% Dec 2025 Official leaderboard
198 Phi-3.5-vision-instruct 63.3% Aug 2024 Official leaderboard
199 Mixtral 8x22B 62.4% Apr 2024 Official leaderboard
200 OLMo 2 32B Instruct 62.4% Mar 2025 Official leaderboard
201 Gemma 3n E4B 62.4% Jun 2025 Official leaderboard
202 Mistral NeMo 61.9% Jul 2024 Official leaderboard
203 Qwen2.5-7B 61.9% Sep 2024 Official leaderboard
204 Phi-3.5-mini 61% Aug 2024 Official leaderboard
205 Amazon Nova Micro 60.5% Dec 2024 Official leaderboard
206 Jamba 1.6 Mini 60.5% Mar 2025 Official leaderboard
207 Gemma 3n E2B 60.5% Jun 2025 Official leaderboard
208 Ministral 8B 59.6% Oct 2024 Official leaderboard
209 Gemma 4 E2B 59.6% Mar 2026 Official leaderboard
210 Ministral 3B 59.2% Oct 2024 Official leaderboard
211 LFM2-8B-A1B 59.2% Oct 2025 Official leaderboard
212 LFM3-1.7B 59.2% Oct 2026 Official leaderboard
213 Phi-4-multimodal 57.8% Feb 2025 Official leaderboard
214 Gemini 1.5 Flash 8B 57.3% Oct 2024 Official leaderboard
215 Jamba 1.5 Mini 56.9% Aug 2024 Official leaderboard
216 Claude 3 Haiku 56% Mar 2024 Official leaderboard
217 Gemma 3 4B 55% Mar 2025 Official leaderboard
218 Jamba 1.7 Mini 54.1% Jul 2025 Official leaderboard
219 Command R7B 53.2% Dec 2024 Official leaderboard
220 GPT-3.5 Turbo Jan 2024 52.3% Jan 2024 Official leaderboard
221 LFM2-1.2B 52.3% Jul 2025 Official leaderboard
222 Qwen3-0.6B 47.3% Apr 2025 Official leaderboard
223 LFM2.5-1.2B-Thinking 45.4% Jan 2026 Official leaderboard
224 Gemma 3 1B 38.1% Mar 2025 Official leaderboard
225 Llama 3.2 1B 28.4% Sep 2024 Official leaderboard
226 SmolLM2-1.7B 28% Oct 2024 Official leaderboard
227 LFM2.5-1.2B-Instruct 28% Jan 2026 Official leaderboard
228 OLMo 2 1B Instruct 26.6% Apr 2025 Official leaderboard
229 SmolLM2-360M 26.1% Oct 2024 Official leaderboard
230 SmolLM2-135M 24.8% Oct 2024 Official leaderboard

More effort adds a median 2 points

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
Gemini 4 Argon
low
89.5%
GPT-6 Astra
medium high
88.1%
GPT-6.1 Terra
low
87.6%
Muse Wonder
low high
86.7%
GPT-6.1 Sol
medium
86.2%
Claude Opus 4.7
Adaptive Th…
85.3%
Claude Opus 5.5
medium
85.3%
Claude Sonnet 5.5
low
84.4%
Grok 4.20
reasoning
83.9%
GPT-5.4
medium
83.9%
Claude Opus 4
Thinking (l…
83.5%
Claude Sonnet 4.5
thinking
83.5%

Source: Official SEND Pedagogy Benchmark leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.1 89.5%
Gemini 4 Argon high 89.5%
Claude Opus 5 89%
Gemini 3.6 Flash 88.5%
Gemini 4 Argon medium 88.5%
GPT-5.5 88.1%
Qwen 3.8 Max 88.1%
Muse Spark 1.2 88.1%
GPT-6 Astra high 88.1%
GPT-5.6 Sol 87.6%
Gemini 4 Argon low 87.6%
GPT-6.1 Terra medium 87.6%
Claude Fable 5 87.2%
Gemini 3.7 Flash 87.2%
Qwen3.8 Max (0902) 87.2%
GPT-6.1 Terra low 87.2%
Gemini 3.1 Pro 86.7%
Muse Wonder high 86.7%
Gemini 3.5 Flash 86.2%
Grok 4.6 86.2%
Qwen3.8-2.4T-A95B 86.2%
Gemini 3.8 Flash 86.2%
GPT-6.1 Sol high 86.2%
Gemini 2.5 Pro Jun 2025 85.8%
Gemini 3 Flash 85.8%
GPT-6.1 Terra high 85.8%
Claude Opus 4.7 85.3%
Qwen3.7-Max 85.3%
GPT-6 Astra medium 85.3%
Claude Opus 5.5 medium 85.3%
GPT-6.1 Sol medium 85.3%
Claude Opus 4.7 Adaptive Thinking 84.9%
Grok 4.5 84.9%
Muse Spark 1.3 84.9%
Gemini 3 Pro 84.4%
Qwen3.8 Flash 84.4%
Grok 4.7 low 84.4%
Claude Sonnet 5.5 medium 84.4%
Muse Wonder medium 84.4%
Grok 4.20 reasoning 83.9%
GPT-5.4 medium 83.9%
Claude Opus 5.5 high 83.9%
Claude Sonnet 5.5 high 83.9%
Claude Sonnet 5.5 low 83.9%
Muse Wonder low 83.9%
Kimi K3.1 Pro 83.9%
Claude Opus 4 Thinking (low) 83.5%
Claude Sonnet 4.5 thinking 83.5%
Qwen3.5 397B-A17B 83.5%
Muse Spark 1.1 83.5%
Kimi K2.6 83%
DeepSeek-V4-Flash 83%
Claude Opus 4.8 83%
Inkling 83%
Tencent Hy4 preview 83%
GLM-5 82.6%
Qwen 3.6 Max (Preview) 82.6%
Claude Sonnet 5 82.6%
o3 82.1%
Grok 4 82.1%
GPT-5.1 medium 82.1%
DeepSeek-V4-Pro 82.1%
Grok 4.3 82.1%
GPT-5.6 Terra 82.1%
Inkling-Small 82.1%
GLM-5.3 82.1%
GPT-5 medium 81.7%
GLM-5.1 81.7%
Kimi K3 81.7%
DeepSeek V4.1 Flash 81.7%
Qwen3-235B-A22B-Thinking Jul 2025 81.2%
GLM-4.5 81.2%
Kimi K2 Thinking 81.2%
Qwen3.5-122B-A10B 81.2%
GLM-5.2 81.2%
Claude Opus 4 80.7%
Claude Sonnet 4 Thinking (low) 80.7%
Kimi K2 Sep 2025 80.7%
Claude Opus 4.5 80.7%
GPT-5.2 medium 80.7%
Claude Opus 4.6 80.7%
Qwen3.5-Plus 80.7%
Claude Sonnet 4.6 Adaptive Thinking 80.7%
Qwen 3.6 Plus 80.7%
Gemma 4 31B IT 80.7%
Nemotron 3 Ultra 80.7%
Gemini 2.5 Flash Sep 2025 80.3%
Grok 4.1 Fast 80.3%
GLM-4.7 80.3%
GPT-5.4 80.3%
GPT-5.6 Luna 80.3%
DeepSeek V4.1 Flash 1017 80.3%
Claude Sonnet 4 79.8%
Gemini 3.1 Flash-Lite 79.8%
o1 medium 79.4%
GPT-4.1 79.4%
GPT-5 mini medium 79.4%
Claude Sonnet 4.5 79.4%
Mistral Large 3 79.4%
Gemini 3.1 Flash-Lite 79.4%
MiMo-V2-Pro 79.4%
MiniMax-M3 79.4%
Muse Glimmer 79.4%
Qwen 3.8 27B 79.4%
Claude Haiku 5.5 medium 79.4%
Claude 3.7 Sonnet Thinking (medium) 78.9%
Mistral Medium 3.1 78.9%
GLM-4.6 78.9%
Qwen3.5-27B 78.9%
MiMo-V2.6 Pro 78.9%
Claude Haiku 5.5 high 78.9%
o4-mini 78.4%
DeepSeek-R1 May 2025 78.4%
Gemini 2.5 Flash Sep 2025 78.4%
Kimi K2.5 78.4%
DeepSeek-V4-Pro 78.4%
MiMo-V2.6 Flash 78.4%
GPT-4.5 78%
Claude Sonnet 4.6 78%
Qwen3.5-35B-A3B 78%
GPT-5.4 Mini medium 78%
Qwen3.6 27B 78%
Mistral Large 4 78%
Grok 4 Fast 77.5%
Qwen3.5-Flash 77.5%
MiniMax-M2.7 77.5%
Qwen 3.6 35B-A3B 77.5%
MiMo-V2.5 77.5%
Qwen3.7 Flash 77.5%
Grok-2 Dec 2024 77.1%
Mistral Medium 3 77.1%
Gemma 4 26B A4B 77.1%
Gemma 4 12B 77.1%
Qwen 3.6 Flash 76.6%
Mistral Medium 3.5 76.6%
Gemini 3.5 Flash-Lite 76.6%
MiniMax-M2.1 76.2%
Gemini 2.0 Flash Feb 2025 75.7%
Grok-3 mini 75.7%
Claude Haiku 4.5 75.7%
MiMo-V2.5-Pro 75.7%
Claude Haiku 5.5 low 75.7%
Llama 4 Maverick 75.2%
GPT-4.1 mini 75.2%
Qwen3-235B-A22B 75.2%
Qwen3-30B-A3B 75.2%
Qwen3-32B 75.2%
Magistral Medium 1.0 74.8%
Nemotron 3.5 Lightning 30B-A3B 74.3%
DeepSeek-V3 Mar 2025 73.8%
GPT-4o Mar 2025 73.8%
gpt-oss-120b 73.8%
Gemma 2 27B 73.4%
Gemma 3 27B 73.4%
Gemini 2.5 Flash-Lite Sep 2025 73.4%
Command A+ 73.4%
Doubao-1.5-pro 72.9%
DeepSeek-V3 Mar 2025 72.9%
LFM2.5-24B-A2B 72.9%
Llama 3.1-405B 72.5%
o3-mini 72.5%
DeepSeek-V3.2 72.5%
Qwen3.5-9B 72.5%
GPT-5.4 Nano medium 72.5%
Gemini 1.5 Pro Sep 2024 72%
Cohere Command A 72%
Grok 3 72%
Gemini 2.0 Flash-Lite 71.6%
Gemini 2.5 Flash-Lite Sep 2025 71.6%
MiniMax-M3.1-Flash 71.6%
learnlm-1.5-pro-experimental 71.1%
GPT-4o mini 71.1%
Mistral Large 2 Nov 2024 71.1%
Amazon Nova Pro 71.1%
Qwen2.5-72B 70.6%
Llama 3.2 90B 70.6%
GPT-5 nano medium 70.6%
Gemma 4 E4B 70.6%
GPT-4 Turbo Apr 2024 70.2%
Llama 3.1-70B 70.2%
Claude 3.7 Sonnet 70.2%
Qwen3-14B 70.2%
gpt-oss-20b 70.2%
DeepSeek-V3.1 70.2%
Llama 3.3 70B 69.7%
Olmo 3.1 32B Think 69.7%
GPT-5.4 Mini 69.7%
Phi-4 69.3%
Mistral Small 3.1 69.3%
Qwen3-8B 69.3%
Mistral Small 3.2 69.3%
GPT-4 Jun 2023 68.8%
Claude 3 Opus 68.3%
Gemini 1.5 Flash Sep 2024 68.3%
o1-mini 68.3%
Amazon Nova Lite 68.3%
Mistral Small 3 68.3%
glm-4.7-flash 68.3%
Llama 4 Scout 67.9%
Nova 2 Lite 67.9%
Jamba 1.5 Large 67.4%
Jamba 1.7 Large 67.4%
Nemotron 3 Nano 30B-A3B 67.4%
MiMo-V2-Flash 67.4%
Granite 4.2 8B 67.4%
Claude 3.5 Haiku 67%
Ministral 3 14B 67%
Claude 3.5 Sonnet 66.1%
Command R+ 66.1%
Claude 3.5 Sonnet (October 2024) 66.1%
Jamba 1.6 Large 66.1%
Qwen3-4B 66.1%
Yi-Large 65.6%
Gemma 3 12B 65.6%
Mistral Small 4 65.6%
Llama 3.2 11B 65.1%
sonar 65.1%
Mixtral 8x7B 64.7%
Llama 3.1-8B 64.7%
GPT-4.1 nano 64.2%
GPT-5.4 Nano 64.2%
Yi-Lightning 63.8%
Qwen3-1.7B 63.8%
Olmo 3.1 32B Instruct 63.8%
Phi-3.5-vision-instruct 63.3%
Grok 4.20 63.3%
Mixtral 8x22B 62.4%
OLMo 2 32B Instruct 62.4%
Gemma 3n E4B 62.4%
Mistral NeMo 61.9%
Qwen2.5-7B 61.9%
Phi-3.5-mini 61%
Amazon Nova Micro 60.5%
Jamba 1.6 Mini 60.5%
Gemma 3n E2B 60.5%
Ministral 8B 59.6%
Gemma 4 E2B 59.6%
Ministral 3B 59.2%
LFM2-8B-A1B 59.2%
LFM3-1.7B 59.2%
Phi-4-multimodal 57.8%
Gemini 1.5 Flash 8B 57.3%
Jamba 1.5 Mini 56.9%
Claude 3 Haiku 56%
Gemma 3 4B 55%
Jamba 1.7 Mini 54.1%
Command R7B 53.2%
GPT-3.5 Turbo Jan 2024 52.3%
LFM2-1.2B 52.3%
Qwen3-0.6B 47.3%
LFM2.5-1.2B-Thinking 45.4%
Gemma 3 1B 38.1%
Llama 3.2 1B 28.4%
SmolLM2-1.7B 28%
LFM2.5-1.2B-Instruct 28%
OLMo 2 1B Instruct 26.6%
SmolLM2-360M 26.1%
SmolLM2-135M 24.8%

More effort adds a median 2 points

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
Gemini 4 Argon
low
89.5%
GPT-6 Astra
medium high
88.1%
GPT-6.1 Terra
low
87.6%
Muse Wonder
low high
86.7%
GPT-6.1 Sol
medium
86.2%
Claude Opus 4.7
Adaptive Th…
85.3%
Claude Opus 5.5
medium
85.3%
Claude Sonnet 5.5
low
84.4%
Grok 4.20
reasoning
83.9%
GPT-5.4
medium
83.9%
Claude Opus 4
Thinking (l…
83.5%
Claude Sonnet 4.5
thinking
83.5%

Source: Official SEND Pedagogy Benchmark leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Claude Fable 5.1 89.5%
Gemini 4 Argon high 89.5%
Claude Opus 5 89%
Gemini 3.6 Flash 88.5%
Gemini 4 Argon medium 88.5%
GPT-5.5 88.1%
Qwen 3.8 Max 88.1%
Muse Spark 1.2 88.1%
GPT-6 Astra high 88.1%
GPT-5.6 Sol 87.6%
Gemini 4 Argon low 87.6%
GPT-6.1 Terra medium 87.6%
Claude Fable 5 87.2%
Gemini 3.7 Flash 87.2%
Qwen3.8 Max (0902) 87.2%
GPT-6.1 Terra low 87.2%
Gemini 3.1 Pro 86.7%
Muse Wonder high 86.7%
Gemini 3.5 Flash 86.2%
Grok 4.6 86.2%
Qwen3.8-2.4T-A95B 86.2%
Gemini 3.8 Flash 86.2%
GPT-6.1 Sol high 86.2%
Gemini 2.5 Pro Jun 2025 85.8%
Gemini 3 Flash 85.8%
GPT-6.1 Terra high 85.8%
Claude Opus 4.7 85.3%
Qwen3.7-Max 85.3%
GPT-6 Astra medium 85.3%
Claude Opus 5.5 medium 85.3%
GPT-6.1 Sol medium 85.3%
Claude Opus 4.7 Adaptive Thinking 84.9%
Grok 4.5 84.9%
Muse Spark 1.3 84.9%
Gemini 3 Pro 84.4%
Qwen3.8 Flash 84.4%
Grok 4.7 low 84.4%
Claude Sonnet 5.5 medium 84.4%
Muse Wonder medium 84.4%
Grok 4.20 reasoning 83.9%
GPT-5.4 medium 83.9%
Claude Opus 5.5 high 83.9%
Claude Sonnet 5.5 high 83.9%
Claude Sonnet 5.5 low 83.9%
Muse Wonder low 83.9%
Kimi K3.1 Pro 83.9%
Claude Opus 4 Thinking (low) 83.5%
Claude Sonnet 4.5 thinking 83.5%
Qwen3.5 397B-A17B 83.5%
Muse Spark 1.1 83.5%
Kimi K2.6 83%
DeepSeek-V4-Flash 83%
Claude Opus 4.8 83%
Inkling 83%
Tencent Hy4 preview 83%
GLM-5 82.6%
Qwen 3.6 Max (Preview) 82.6%
Claude Sonnet 5 82.6%
o3 82.1%
Grok 4 82.1%
GPT-5.1 medium 82.1%
DeepSeek-V4-Pro 82.1%
Grok 4.3 82.1%
GPT-5.6 Terra 82.1%
Inkling-Small 82.1%
GLM-5.3 82.1%
GPT-5 medium 81.7%
GLM-5.1 81.7%
Kimi K3 81.7%
DeepSeek V4.1 Flash 81.7%
Qwen3-235B-A22B-Thinking Jul 2025 81.2%
GLM-4.5 81.2%
Kimi K2 Thinking 81.2%
Qwen3.5-122B-A10B 81.2%
GLM-5.2 81.2%
Claude Opus 4 80.7%
Claude Sonnet 4 Thinking (low) 80.7%
Kimi K2 Sep 2025 80.7%
Claude Opus 4.5 80.7%
GPT-5.2 medium 80.7%
Claude Opus 4.6 80.7%
Qwen3.5-Plus 80.7%
Claude Sonnet 4.6 Adaptive Thinking 80.7%
Qwen 3.6 Plus 80.7%
Gemma 4 31B IT 80.7%
Nemotron 3 Ultra 80.7%
Gemini 2.5 Flash Sep 2025 80.3%
Grok 4.1 Fast 80.3%
GLM-4.7 80.3%
GPT-5.4 80.3%
GPT-5.6 Luna 80.3%
DeepSeek V4.1 Flash 1017 80.3%
Claude Sonnet 4 79.8%
Gemini 3.1 Flash-Lite 79.8%
o1 medium 79.4%
GPT-4.1 79.4%
GPT-5 mini medium 79.4%
Claude Sonnet 4.5 79.4%
Mistral Large 3 79.4%
Gemini 3.1 Flash-Lite 79.4%
MiMo-V2-Pro 79.4%
MiniMax-M3 79.4%
Muse Glimmer 79.4%
Qwen 3.8 27B 79.4%
Claude Haiku 5.5 medium 79.4%
Claude 3.7 Sonnet Thinking (medium) 78.9%
Mistral Medium 3.1 78.9%
GLM-4.6 78.9%
Qwen3.5-27B 78.9%
MiMo-V2.6 Pro 78.9%
Claude Haiku 5.5 high 78.9%
o4-mini 78.4%
DeepSeek-R1 May 2025 78.4%
Gemini 2.5 Flash Sep 2025 78.4%
Kimi K2.5 78.4%
DeepSeek-V4-Pro 78.4%
MiMo-V2.6 Flash 78.4%
GPT-4.5 78%
Claude Sonnet 4.6 78%
Qwen3.5-35B-A3B 78%
GPT-5.4 Mini medium 78%
Qwen3.6 27B 78%
Mistral Large 4 78%
Grok 4 Fast 77.5%
Qwen3.5-Flash 77.5%
MiniMax-M2.7 77.5%
Qwen 3.6 35B-A3B 77.5%
MiMo-V2.5 77.5%
Qwen3.7 Flash 77.5%
Grok-2 Dec 2024 77.1%
Mistral Medium 3 77.1%
Gemma 4 26B A4B 77.1%
Gemma 4 12B 77.1%
Qwen 3.6 Flash 76.6%
Mistral Medium 3.5 76.6%
Gemini 3.5 Flash-Lite 76.6%
MiniMax-M2.1 76.2%
Gemini 2.0 Flash Feb 2025 75.7%
Grok-3 mini 75.7%
Claude Haiku 4.5 75.7%
MiMo-V2.5-Pro 75.7%
Claude Haiku 5.5 low 75.7%
Llama 4 Maverick 75.2%
GPT-4.1 mini 75.2%
Qwen3-235B-A22B 75.2%
Qwen3-30B-A3B 75.2%
Qwen3-32B 75.2%
Magistral Medium 1.0 74.8%
Nemotron 3.5 Lightning 30B-A3B 74.3%
DeepSeek-V3 Mar 2025 73.8%
GPT-4o Mar 2025 73.8%
gpt-oss-120b 73.8%
Gemma 2 27B 73.4%
Gemma 3 27B 73.4%
Gemini 2.5 Flash-Lite Sep 2025 73.4%
Command A+ 73.4%
Doubao-1.5-pro 72.9%
DeepSeek-V3 Mar 2025 72.9%
LFM2.5-24B-A2B 72.9%
Llama 3.1-405B 72.5%
o3-mini 72.5%
DeepSeek-V3.2 72.5%
Qwen3.5-9B 72.5%
GPT-5.4 Nano medium 72.5%
Gemini 1.5 Pro Sep 2024 72%
Cohere Command A 72%
Grok 3 72%
Gemini 2.0 Flash-Lite 71.6%
Gemini 2.5 Flash-Lite Sep 2025 71.6%
MiniMax-M3.1-Flash 71.6%
learnlm-1.5-pro-experimental 71.1%
GPT-4o mini 71.1%
Mistral Large 2 Nov 2024 71.1%
Amazon Nova Pro 71.1%
Qwen2.5-72B 70.6%
Llama 3.2 90B 70.6%
GPT-5 nano medium 70.6%
Gemma 4 E4B 70.6%
GPT-4 Turbo Apr 2024 70.2%
Llama 3.1-70B 70.2%
Claude 3.7 Sonnet 70.2%
Qwen3-14B 70.2%
gpt-oss-20b 70.2%
DeepSeek-V3.1 70.2%
Llama 3.3 70B 69.7%
Olmo 3.1 32B Think 69.7%
GPT-5.4 Mini 69.7%
Phi-4 69.3%
Mistral Small 3.1 69.3%
Qwen3-8B 69.3%
Mistral Small 3.2 69.3%
GPT-4 Jun 2023 68.8%
Claude 3 Opus 68.3%
Gemini 1.5 Flash Sep 2024 68.3%
o1-mini 68.3%
Amazon Nova Lite 68.3%
Mistral Small 3 68.3%
glm-4.7-flash 68.3%
Llama 4 Scout 67.9%
Nova 2 Lite 67.9%
Jamba 1.5 Large 67.4%
Jamba 1.7 Large 67.4%
Nemotron 3 Nano 30B-A3B 67.4%
MiMo-V2-Flash 67.4%
Granite 4.2 8B 67.4%
Claude 3.5 Haiku 67%
Ministral 3 14B 67%
Claude 3.5 Sonnet 66.1%
Command R+ 66.1%
Claude 3.5 Sonnet (October 2024) 66.1%
Jamba 1.6 Large 66.1%
Qwen3-4B 66.1%
Yi-Large 65.6%
Gemma 3 12B 65.6%
Mistral Small 4 65.6%
Llama 3.2 11B 65.1%
sonar 65.1%
Mixtral 8x7B 64.7%
Llama 3.1-8B 64.7%
GPT-4.1 nano 64.2%
GPT-5.4 Nano 64.2%
Yi-Lightning 63.8%
Qwen3-1.7B 63.8%
Olmo 3.1 32B Instruct 63.8%
Phi-3.5-vision-instruct 63.3%
Grok 4.20 63.3%
Mixtral 8x22B 62.4%
OLMo 2 32B Instruct 62.4%
Gemma 3n E4B 62.4%
Mistral NeMo 61.9%
Qwen2.5-7B 61.9%
Phi-3.5-mini 61%
Amazon Nova Micro 60.5%
Jamba 1.6 Mini 60.5%
Gemma 3n E2B 60.5%
Ministral 8B 59.6%
Gemma 4 E2B 59.6%
Ministral 3B 59.2%
LFM2-8B-A1B 59.2%
LFM3-1.7B 59.2%
Phi-4-multimodal 57.8%
Gemini 1.5 Flash 8B 57.3%
Jamba 1.5 Mini 56.9%
Claude 3 Haiku 56%
Gemma 3 4B 55%
Jamba 1.7 Mini 54.1%
Command R7B 53.2%
GPT-3.5 Turbo Jan 2024 52.3%
LFM2-1.2B 52.3%
Qwen3-0.6B 47.3%
LFM2.5-1.2B-Thinking 45.4%
Gemma 3 1B 38.1%
Llama 3.2 1B 28.4%
SmolLM2-1.7B 28%
LFM2.5-1.2B-Instruct 28%
OLMo 2 1B Instruct 26.6%
SmolLM2-360M 26.1%
SmolLM2-135M 24.8%

9 models have held number one on the official leaderboard since its release in March 2025

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

OpenAI Anthropic Google DeepMind
Own leaderboard Passed 85% in Jun 2025, by Gemini 2.5 Pro

Source: Official SEND Pedagogy Benchmark leaderboard

Gemini 2.5 Pro on the official leaderboard passed 85% in June 2025

That was 3 months after its release. By the same age, 12% of the benchmarks from 2025 had passed 85%. 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 Best so far
0% 20% 40% 60% 80% 100% 2023 2024 2025 2026 2027 Near the ceiling (85%) Claude Fable 5.1 Gemini 2.5 Pro o1 Gemma 2 27B GPT-4 0% 20% 40% 60% 80% 100% 2023 2024 2025 2026 2027 Near the ceiling (85%) Claude Fable 5.1 o3

Source: Official SEND Pedagogy Benchmark leaderboard

No lab model card reports it yet

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

Versions

More in Knowledge

All 332 in Knowledge
  • MMLU-Pro In use Harder 10-option version of MMLU with reasoning-focused questions across 14 subjects. Intern-S2-397B 89.8%
  • MMLU In use 57-subject multiple-choice exam questions from elementary to professional level. Scored by accuracy. Seed1.8 92.3%
  • SuperGPQA In use 26,529 graduate-level multiple-choice questions across 285 disciplines. Accuracy. Qwen 3.6 Max (Preview) 73.9%
  • WideSearch In use 200 bilingual broad info-seeking tasks: agent collects many atomic facts into a table. Scored by success rate and row/item F1. Tencent Hy4 preview 83.9%
  • BBEH micro average In use Models solve harder replacements for each of the 23 BIG-Bench Hard tasks, 4,520 examples in total. The score is the share of all examples answered correctly. Gemma 4 31B IT 74.4%
  • HLE-Diamond In use Models answer 1,000 cleaned expert questions from Humanity's Last Exam, 500 on reasoning and 500 on knowledge, closed book. Graded on exact answers. Claude Fable 5.2 77.9%

Data audit

All 10 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 (247 entries).
  • Pass Models scored 230 models we track, plus 17 other entries on its leaderboard.
  • Pass Newest score The newest is from October 2026, for LFM3-1.7B.
  • Pass Frontier models 7 of the top 10 have a score, led by GPT-6 Astra (#4).
  • Note Headroom Saturated. The best is 89.5%, past 85% since June 2025.
Leaderboard import 3 pass
  • Pass Importer We read the board on fab-ai.org: 258 rows for models we track and 17 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check