Benchmark Almanac

LMArena (Elo)

Gemini 4 Carbon 1,528 leads since Oct 2026 Fading

People chat with two anonymous models side by side and vote for the better reply. Votes become Elo-style ratings for each model.

The 2025 paper The Leaderboard Illusion reported that some labs tested many private variants before release. Arena changes vote filters and the rating fit over time, for example two bias terms added on 2026-05-12, so old and new ratings drift. Arena Expert, added 2025-11-05, ranks the hardest 5.5% of prompts.

In the authors’ words

Large Language Models (LLMs) have unlocked new capabilities and applications; however, evaluating the alignment with human preferences still poses significant challenges. To address this issue, we introduce Chatbot Arena, an open platform for evaluating LLMs based on human preferences. Our methodology employs a pairwise comparison approach and leverages input from a diverse user base through crowdsourcing. The platform has been operational for several months, amassing over 240K votes. This paper describes the platform, analyzes the data we have collected so far, and explains the tried-and-true statistical methods we are using for efficient and accurate evaluation and ranking of models. We confirm that the crowdsourced questions are sufficiently diverse and discriminating and that the crowdsourced human votes are in good agreement with those of expert raters. These analyses collectively establish a robust foundation for the credibility of Chatbot Arena. Because of its unique value and openness, Chatbot Arena has emerged as one of the most referenced LLM leaderboards, widely cited by leading LLM developers and companies. Our demo is publicly available at https://chat.lmsys.org.

Abstract of Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference (arXiv:2403.04132, CC0)
Made by Arena (formerly LMArena) Released May 2023 Overall chat quality Last in a launch post Dec 2025 , dropped before it saturated Last model added Oct 2026 Scores refreshed 2 Nov 2026
Scored
Ranked by votes People pick the better of two answers, as in an arena, and the votes are turned into ratings.
Answers
Open-ended Free-form text, documents, designs or media judged as a whole.
lm-sys/FastChat 39,552 stars Paper Dataset Leaderboard arena.ai
383
Entries on its own leaderboard, 333 models we track
5
Launch posts list it, from 3 labs
13
System cards report it, from 5 labs

Gemini 4 Carbon leads with 1,528

Best score per entry from the benchmark's own leaderboard, read on 2 Nov 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.

383 entries. Scroll sideways for the rest. Bars start at 800.

  1. 1,528 Gemini 4 Carbon
  2. 1,520 Gemini 4 Argon
  3. 1,509 Claude Opus 5.5
  4. 1,507 Muse Wonder
  5. 1,504 Claude Fable 5.1
  6. 1,504 Claude Fable 5
  7. 1,503 Claude Opus 4.6
  8. 1,501 Claude Opus 4.7
  9. 1,496 Gemini 3.8 Flash
  10. 1,495 Muse Spark 1.3
  11. 1,492 Muse Spark 1.2
  12. 1,492 Claude Opus 5
  13. 1,490 Muse Spark 1.1
  14. 1,490 Gemini 3.7 Flash
  15. 1,489 Muse Spark
  16. 1,488 Kimi K3
  17. 1,487 Gemini 3.1 Pro
  18. 1,485 Gemini 3 Pro
  19. 1,485 GPT-5.6 Sol
  20. 1,483 Gemini 3.6 Flash
  21. 1,482 Qwen 3.8 Max
  22. 1,482 Claude Opus 4.8
  23. 1,482 GPT-5.5
  24. 1,481 GPT-6.1 Sol
  25. 1,481 GLM-5.3
  26. 1,480 MiMo-V2.6 Pro
  27. 1,478 Gemini 3.5 Flash
  28. 1,477 Claude Sonnet 5.5
  29. 1,477 GLM-5.3-Flash
  30. 1,476 DeepSeek V4.1 Flash
  31. 1,476 GPT-5.2 Instant
  32. 1,476 GPT-6 Astra
  33. 1,475 GLM-5.2
  34. 1,475 GPT-5.4
  35. 1,475 Grok 4.20 Beta 1
  36. 1,474 Qwen3.7-Max
  37. 1,474 Claude Opus 4.5
  38. 1,473 GPT-5.5 Instant
  39. 1,473 Gemini 3 Flash
  40. 1,472 Claude Sonnet 4.6
  41. 1,472 Grok 4.20 Beta (Reasoning, 0309)
  42. 1,471 Grok 4.20 Multi-Agent Beta (0309)
  43. 1,467 GPT-5.6 Terra
  44. 1,467 ERNIE 5.1
  45. 1,467 MiMo-V2.5-Pro
  46. 1,465 Grok 4.5
  47. 1,465 GLM-5.1
  48. 1,465 DeepSeek-V4-Pro
  49. 1,465 Grok 4.1
  50. 1,465 Qwen3.5-Max-Preview
  51. 1,462 Claude Sonnet 5
  52. 1,461 Kimi K2.6
  53. 1,460 Qwen 3.6 Max (Preview)
  54. 1,458 GPT-6 Sol
  55. 1,457 GLM-5
  56. 1,456 Claude Sonnet 4.5
  57. 1,456 Seed2.0 Pro
  58. 1,456 GPT-5.1
  59. 1,456 Qwen3.7-Plus
  60. 1,456 Gemini 3.5 Flash-Lite
  61. 1,455 Grok 4.6
  62. 1,454 Tencent Hy3
  63. 1,453 GPT-5.6 Luna
  64. 1,452 Gemma 4 31B IT
  65. 1,450 GPT-5.3 Instant
  66. 1,450 Claude Opus 4.1
  67. 1,450 Kimi K2.5
  68. 1,448 ERNIE-5.0-Preview (1203)
  69. 1,448 Step 5 Preview
  70. 1,448 MiMo-V2-Pro
  71. 1,448 MiMo-V2.6 Flash
  72. 1,447 GPT-5.4 Mini
  73. 1,446 ERNIE 5.0 (0110)
  74. 1,445 Gemini 2.5 Pro
  75. 1,445 GPT-4.5
  76. 1,443 Qwen 3.6 Plus
  77. 1,443 GPT-4o
  78. 1,442 Inkling
  79. 1,442 Grok 4.3
  80. 1,441 Qwen3.5 397B-A17B
  81. 1,441 GLM-4.7
  82. 1,441 Grok 4.7
  83. 1,441 GPT-6 Luna
  84. 1,439 MiniMax-M3
  85. 1,439 DeepSeek-V4-Flash
  86. 1,438 Gemma 4 26B A4B
  87. 1,437 GPT-5.2
  88. 1,436 LongCat-Flash-Chat Exp (2602)
  89. 1,436 Beam 1.0
  90. 1,436 Qwen 3.8 27B
  91. 1,435 GPT-5
  92. 1,434 Qwen3-Max
  93. 1,434 GLM-5V-Turbo
  94. 1,434 MiMo-V2.5
  95. 1,433 Gemini 3.1 Flash-Lite
  96. 1,432 Mistral Large 4
  97. 1,432 o3
  98. 1,431 MiMo-V2-Omni
  99. 1,430 Grok 4.1 Fast
  100. 1,430 Kimi K2 Thinking Turbo
  101. 1,427 amazon-nova-experimental-chat (26-02-10)
  102. 1,427 Nemotron 3 Ultra
  103. 1,426 GPT-5 Chat
  104. 1,426 Mistral Medium 3.5
  105. 1,426 Claude Opus 4
  106. 1,425 DeepSeek-V3.2-Exp
  107. 1,425 Muse Glimmer
  108. 1,425 Claude Haiku 5.5
  109. 1,425 DeepSeek-V3.2
  110. 1,424 GLM-4.6
  111. 1,422 Qwen3-235B-A22B-Instruct
  112. 1,421 DeepSeek-R1
  113. 1,418 Kimi K2
  114. 1,418 Grok 4 Fast
  115. 1,418 ERNIE-5.0-Preview (1022)
  116. 1,418 Kimi K2
  117. 1,417 DeepSeek-V3.1
  118. 1,416 Qwen3.5-122B-A10B
  119. 1,416 DeepSeek-V3.1-Terminus
  120. 1,415 MiniMax-M2.7
  121. 1,415 GPT-4.1
  122. 1,415 Claude Haiku 4.5
  123. 1,414 Mistral Large 3
  124. 1,413 Qwen3-VL-235B-A22B-Instruct
  125. 1,413 amazon-nova-experimental-chat (26-01-10)
  126. 1,412 Nemotron 3.5 Super VL
  127. 1,411 Grok 3
  128. 1,411 Grok 4
  129. 1,411 GLM-4.5
  130. 1,409 Gemini 2.5 Flash
  131. 1,409 Qwen3.5-27B
  132. 1,409 Tencent Hy3 preview
  133. 1,408 Mistral Medium 3.1
  134. 1,406 Inkling-Small
  135. 1,403 Gemini 2.5 Flash
  136. 1,402 o1
  137. 1,402 Qwen3-235B-A22B
  138. 1,402 Claude Sonnet 4
  139. 1,401 GPT-5.4 Nano
  140. 1,401 LongCat-Flash-Chat
  141. 1,400 Qwen3-235B-A22B-Thinking
  142. 1,399 Qwen3-Next-80B-A3B-Instruct
  143. 1,398 DeepSeek-R1
  144. 1,396 Qwen3.5-Flash
  145. 1,395 DeepSeek-V3
  146. 1,395 Qwen3-VL-235B-A22B
  147. 1,394 Hunyuan Vision 1.5 Thinking
  148. 1,394 Qwen3.5-35B-A3B
  149. 1,393 Step 3.5 Flash
  150. 1,393 amazon-nova-experimental-chat (12-10)
  151. 1,392 MiMo-V2-Flash
  152. 1,391 o4-mini
  153. 1,391 MiniMax-M2.5
  154. 1,390 GPT-5 mini
  155. 1,389 o1-preview
  156. 1,388 Claude 3.7 Sonnet
  157. 1,388 Solar Pro 4
  158. 1,388 Qwen3-Coder-480B-A35B
  159. 1,387 Hunyuan-T1
  160. 1,387 Mistral Medium 3
  161. 1,385 MiniMax-M2.1
  162. 1,383 Hunyuan-TurboS
  163. 1,383 GPT-4.1 mini
  164. 1,382 Qwen3-30B-A3B-Instruct
  165. 1,379 Gemini 2.5 Flash-Lite
  166. 1,379 Trinity Large Preview
  167. 1,378 GLM-4.6V
  168. 1,375 Gemini 2.5 Flash-Lite
  169. 1,374 Claude 3.5 Sonnet (October 2024)
  170. 1,374 Qwen2.5-Max
  171. 1,373 GLM-4.5-Air
  172. 1,369 Qwen3-Next-80B-A3B
  173. 1,367 Trinity Large Thinking
  174. 1,365 Gemma 3 27B
  175. 1,365 glm-4.7-flash
  176. 1,364 MiniMax-M1-80k
  177. 1,364 amazon-nova-experimental-chat (11-10)
  178. 1,364 Grok-3 mini
  179. 1,364 o3-mini
  180. 1,360 Gemini 2.0 Flash
  181. 1,360 Nemotron 3 Super
  182. 1,358 DeepSeek-V3
  183. 1,356 Mistral Small 3.2
  184. 1,356 INTELLECT-3
  185. 1,354 Cohere Command A
  186. 1,354 Gemini 2.0 Flash-Lite
  187. 1,353 Nemotron 3.5 Lightning 30B-A3B
  188. 1,352 GLM-4.5V
  189. 1,352 gpt-oss-120b
  190. 1,351 Gemini 1.5 Pro
  191. 1,349 Step 3
  192. 1,348 amazon-nova-experimental-chat (10-20)
  193. 1,348 Llama-3.1-Nemotron-Ultra-253B-v1
  194. 1,347 Qwen3-32B
  195. 1,346 amazon-nova-experimental-chat (10-09)
  196. 1,346 GPT-4o
  197. 1,346 Qwen-Plus (0125)
  198. 1,344 MiniMax-M2
  199. 1,343 Claude 3.5 Sonnet
  200. 1,343 Ling-flash-2.0
  201. 1,343 Llama-3.3-Nemotron-Super-49B-v1.5
  202. 1,343 Mercury 2
  203. 1,343 GLM-4-Plus (0111)
  204. 1,342 Gemma 3 12B
  205. 1,341 Hunyuan Turbo (0110)
  206. 1,340 Granite 4.2 30B
  207. 1,338 GPT-5 nano
  208. 1,337 o1-mini
  209. 1,336 Gemini Advanced (0514)
  210. 1,336 Grok-2
  211. 1,336 GPT-4o
  212. 1,336 QwQ-32B
  213. 1,335 Llama 3.1-405B
  214. 1,335 Nova 2 Lite
  215. 1,334 Step-2 16K Exp (202412)
  216. 1,329 Olmo 3.1 32B Instruct
  217. 1,328 Yi-Lightning
  218. 1,328 Llama-3.3-Nemotron-Super-49B-v1
  219. 1,327 Llama 4 Maverick
  220. 1,327 Qwen3-30B-A3B
  221. 1,326 Hunyuan Large 20250210
  222. 1,325 Claude 3.5 Haiku
  223. 1,325 GPT-4 Turbo
  224. 1,324 Gemini 1.5 Pro
  225. 1,323 DeepSeek-V2.5-1210
  226. 1,322 Ring-flash-2.0
  227. 1,322 GPT-4.1 nano
  228. 1,322 Claude 3 Opus
  229. 1,321 Llama 4 Scout
  230. 1,321 Molmo2-8B
  231. 1,320 Step-1o Turbo (202506)
  232. 1,319 GLM-4-Plus
  233. 1,318 Qwen-Max (0919)
  234. 1,318 Llama 3.3 70B Instruct
  235. 1,318 GPT-4o mini
  236. 1,318 Gemma 3n E4B
  237. 1,317 gpt-oss-20b
  238. 1,315 Qwen2.5-Plus (1127)
  239. 1,314 Mistral Large 2
  240. 1,314 Athene-V2-Chat
  241. 1,314 Nemotron 3 Nano 30B-A3B
  242. 1,313 GPT-4 Turbo (Jan 2024)
  243. 1,313 GPT-4 Turbo
  244. 1,311 Hunyuan Standard 20250210
  245. 1,309 Gemini 1.5 Flash
  246. 1,308 Grok-2 mini
  247. 1,307 DeepSeek-V2.5
  248. 1,307 Athene-70B
  249. 1,306 Olmo 3 32B Think
  250. 1,306 Mercury
  251. 1,306 Mistral Large 2
  252. 1,305 Magistral Medium 1.0
  253. 1,304 Granite 4.1 8B
  254. 1,303 Gemma 3 4B
  255. 1,303 Qwen2.5-72B
  256. 1,303 Mistral Small 3.1
  257. 1,299 Llama-3.1-Nemotron-70B-Instruct
  258. 1,294 Hunyuan Large Vision
  259. 1,293 Llama 3.1-70B
  260. 1,290 Amazon Nova Pro
  261. 1,290 Granite 4.2 3B
  262. 1,290 Granite 4.2 8B
  263. 1,289 Gemma 2 27B
  264. 1,289 Jamba 1.5 Large
  265. 1,288 Reka Core
  266. 1,288 GPT-4
  267. 1,287 Granite 4.0 H Small
  268. 1,287 Llama-3.1-Nemotron-51B-Instruct
  269. 1,287 Gemini 1.5 Flash
  270. 1,286 Olmo 3.1 32B Think
  271. 1,286 Tulu 3 (Tülu 3) 70B
  272. 1,281 Claude 3 Sonnet
  273. 1,280 Gemma 2 9B IT SimPO
  274. 1,277 Nemotron-4-340B-Instruct
  275. 1,277 Llama 3-70B
  276. 1,276 Command R+
  277. 1,276 GPT-4
  278. 1,274 Mistral Small 3
  279. 1,273 GLM-4 (0520)
  280. 1,272 Reka Flash
  281. 1,271 Qwen2.5-Coder-32B-Instruct
  282. 1,267 Aya Expanse 32B
  283. 1,267 Gemma 2 9B
  284. 1,265 DeepSeek-Coder-V2 236B
  285. 1,262 Qwen2-72B
  286. 1,262 Claude 3 Haiku
  287. 1,260 Amazon Nova Lite
  288. 1,259 Gemini 1.5 Flash 8B
  289. 1,256 Phi-4
  290. 1,252 OLMo 2 32B Instruct
  291. 1,250 c4ai-command-r-08-2024
  292. 1,243 Mistral Large
  293. 1,241 Amazon Nova Micro
  294. 1,240 Jamba 1.5 Mini
  295. 1,238 Ministral 8B
  296. 1,237 Gemini 1.0 Pro
  297. 1,234 Qwen1.5-110B
  298. 1,234 Reka Flash 21B (Online)
  299. 1,234 Qwen1.5-72B
  300. 1,233 Hunyuan Standard 256K
  301. 1,230 Mixtral 8x22B
  302. 1,227 Command R (03-2024)
  303. 1,226 GPT-3.5 Turbo
  304. 1,224 Llama 3-8B
  305. 1,223 Aya Expanse 8B
  306. 1,223 Mistral Medium
  307. 1,221 Llama-3.1-Tulu-3-8B
  308. 1,213 Zephyr ORPO 141B A35B v0.1
  309. 1,213 Yi-1.5-34B
  310. 1,211 Llama 3.1-8B
  311. 1,209 Granite-3.1-8B-Instruct
  312. 1,204 GPT-3.5 Turbo
  313. 1,204 Qwen1.5-32B
  314. 1,200 Gemma 2 2B IT
  315. 1,198 phi-3-medium 14B
  316. 1,197 Mixtral 8x7B
  317. 1,196 DBRX
  318. 1,191 Qwen1.5-14B
  319. 1,191 InternLM2.5-20B-Chat
  320. 1,185 DeepSeek LLM 67B
  321. 1,185 WizardLM 70B
  322. 1,184 Yi-34B
  323. 1,183 Granite-3.0-8B-Instruct
  324. 1,183 OpenChat 3.5
  325. 1,183 OpenChat 3.5 (0106)
  326. 1,183 Gemma 1.1 7B IT
  327. 1,180 Snowflake Arctic Instruct
  328. 1,179 Granite-3.1-2B-Instruct
  329. 1,178 Tulu 2 DPO 70B
  330. 1,176 OpenHermes 2.5 Mistral 7B
  331. 1,173 Vicuna 33B
  332. 1,171 phi-3-small 7.4B
  333. 1,171 Starling-LM-7B-beta
  334. 1,171 Llama 2-70B
  335. 1,167 Starling-LM-7B-alpha
  336. 1,167 Llama 3.2 3B
  337. 1,164 Nous-Hermes-2-Mixtral-8x7B-DPO
  338. 1,157 Granite-3.0-2B-Instruct
  339. 1,155 Llama2-70B-SteerLM-Chat
  340. 1,154 QwQ-32B-Preview
  341. 1,152 SOLAR-10.7B-Instruct-v1.0
  342. 1,152 Dolphin 2.2.1 Mistral 7B
  343. 1,151 MPT-30B
  344. 1,150 WizardLM 13B
  345. 1,149 Mistral 7B v0.2
  346. 1,148 Falcon-180B
  347. 1,144 Qwen1.5-7B
  348. 1,143 Phi-3 Mini 4K Instruct (June 2024)
  349. 1,142 Vicuna 13B
  350. 1,141 Llama 2-13B
  351. 1,139 Qwen-14B
  352. 1,139 PaLM 2
  353. 1,138 Gemma 7B
  354. 1,137 Code Llama 34B Instruct
  355. 1,131 Zephyr 7B Beta
  356. 1,130 Phi-3-mini-128k-instruct
  357. 1,128 phi-3-mini 3.8B
  358. 1,128 Guanaco 33B
  359. 1,127 Zephyr 7B Alpha
  360. 1,122 StripedHyena Nous 7B
  361. 1,119 Code Llama 70B Instruct
  362. 1,117 Gemma 1.1 2B IT
  363. 1,115 Vicuna 7B
  364. 1,115 SmolLM2 1.7B Instruct
  365. 1,111 Llama 3.2 1B
  366. 1,110 Mistral-7B-Instruct-v0.1
  367. 1,108 Llama 2-7B
  368. 1,094 Gemma 2B
  369. 1,091 Qwen1.5-4B-Chat
  370. 1,074 OLMo-7B-Instruct
  371. 1,071 Koala 13B
  372. 1,070 Alpaca 13B
  373. 1,068 GPT4All 13B Snoozy
  374. 1,064 MPT-7B
  375. 1,057 ChatGLM3-6B
  376. 1,042 RWKV-4 Raven 14B
  377. 1,025 chatglm2-6b
  378. 1,024 OpenAssistant Pythia 12B
  379. 996 ChatGLM-6B
  380. 993 FastChat-T5-3B
  381. 983 Dolly 2.0-12b
  382. 975 LLaMA-13B
  383. 953 stablelm-tuned-alpha-7b

Bars start at 800.

  1. 1 Gemini 4 Carbon high 1,528
  2. 2 Gemini 4 Argon high 1,520
  3. 3 Claude Opus 5.5 high 1,509
  4. 4 Muse Wonder max 1,507
  5. 5 Claude Fable 5.1 max 1,504
  6. 6 Claude Fable 5 high 1,504
  7. 7 Claude Opus 4.6 high 1,503
  8. 8 Claude Opus 4.7 high 1,501
  9. 9 Gemini 3.8 Flash high 1,496
  10. 10 Muse Spark 1.3 max 1,495
  11. 11 Muse Spark 1.2 xhigh 1,492
  12. 12 Claude Opus 5 high 1,492
  13. 13 Muse Spark 1.1 1,490
  14. 14 Gemini 3.7 Flash high 1,490
  15. 15 Muse Spark 1,489
  16. 16 Kimi K3 max 1,488
  17. 17 Gemini 3.1 Pro 1,487
  18. 18 Gemini 3 Pro 1,485
  19. 19 GPT-5.6 Sol xhigh 1,485
  20. 20 Gemini 3.6 Flash high 1,483
  21. 21 Qwen 3.8 Max 1,482
  22. 22 Claude Opus 4.8 high 1,482
  23. 23 GPT-5.5 high 1,482
  24. 24 GPT-6.1 Sol max 1,481
  25. 25 GLM-5.3 max 1,481
  26. 26 MiMo-V2.6 Pro 1,480
  27. 27 Gemini 3.5 Flash high 1,478
  28. 28 Claude Sonnet 5.5 xhigh 1,477
  29. 29 GLM-5.3-Flash 1,477
  30. 30 DeepSeek V4.1 Flash max 1,476
  31. 31 GPT-5.2 Instant 1,476
  32. 32 GPT-6 Astra max 1,476
  33. 33 GLM-5.2 max 1,475
  34. 34 GPT-5.4 high 1,475
  35. 35 Grok 4.20 Beta 1 1,475
  36. 36 Qwen3.7-Max 1,474
  37. 37 Claude Opus 4.5 high, 32k 1,474
  38. 38 GPT-5.5 Instant 1,473
  39. 39 Gemini 3 Flash 1,473
  40. 40 Claude Sonnet 4.6 1,472
  41. 41 Grok 4.20 Beta (Reasoning, 0309) 1,472
  42. 42 Grok 4.20 Multi-Agent Beta (0309) 1,471
  43. 43 GPT-5.6 Terra xhigh 1,467
  44. 44 ERNIE 5.1 1,467
  45. 45 MiMo-V2.5-Pro 1,467
  46. 46 Grok 4.5 1,465
  47. 47 GLM-5.1 1,465
  48. 48 DeepSeek-V4-Pro high 1,465
  49. 49 Grok 4.1 thinking 1,465
  50. 50 Qwen3.5-Max-Preview 1,465
  51. 51 Claude Sonnet 5 high 1,462
  52. 52 Kimi K2.6 1,461
  53. 53 Qwen 3.6 Max (Preview) 1,460
  54. 54 GPT-6 Sol max 1,458
  55. 55 GLM-5 1,457
  56. 56 Claude Sonnet 4.5 high, 32k 1,456
  57. 57 Seed2.0 Pro 1,456
  58. 58 GPT-5.1 high 1,456
  59. 59 Qwen3.7-Plus 1,456
  60. 60 Gemini 3.5 Flash-Lite 1,456
  61. 61 Grok 4.6 high 1,455
  62. 62 Tencent Hy3 1,454
  63. 63 GPT-5.6 Luna xhigh 1,453
  64. 64 Gemma 4 31B IT 1,452
  65. 65 GPT-5.3 Instant 1,450
  66. 66 Claude Opus 4.1 thinking 16k 1,450
  67. 67 Kimi K2.5 thinking 1,450
  68. 68 ERNIE-5.0-Preview (1203) 1,448
  69. 69 Step 5 Preview 1,448
  70. 70 MiMo-V2-Pro 1,448
  71. 71 MiMo-V2.6 Flash 1,448
  72. 72 GPT-5.4 Mini high 1,447
  73. 73 ERNIE 5.0 (0110) 1,446
  74. 74 Gemini 2.5 Pro Jun 2025 1,445
  75. 75 GPT-4.5 1,445
  76. 76 Qwen 3.6 Plus 1,443
  77. 77 GPT-4o Mar 2025 1,443
  78. 78 Inkling 1,442
  79. 79 Grok 4.3 1,442
  80. 80 Qwen3.5 397B-A17B 1,441
  81. 81 GLM-4.7 1,441
  82. 82 Grok 4.7 xhigh 1,441
  83. 83 GPT-6 Luna max 1,441
  84. 84 MiniMax-M3 1,439
  85. 85 DeepSeek-V4-Flash high 1,439
  86. 86 Gemma 4 26B A4B 1,438
  87. 87 GPT-5.2 high 1,437
  88. 88 LongCat-Flash-Chat Exp (2602) 1,436
  89. 89 Beam 1.0 1,436
  90. 90 Qwen 3.8 27B 1,436
  91. 91 GPT-5 high 1,435
  92. 92 Qwen3-Max 1,434
  93. 93 GLM-5V-Turbo 1,434
  94. 94 MiMo-V2.5 1,434
  95. 95 Gemini 3.1 Flash-Lite 1,433
  96. 96 Mistral Large 4 1,432
  97. 97 o3 1,432
  98. 98 MiMo-V2-Omni 1,431
  99. 99 Grok 4.1 Fast reasoning 1,430
  100. 100 Kimi K2 Thinking Turbo 1,430
  101. 101 amazon-nova-experimental-chat (26-02-10) 1,427
  102. 102 Nemotron 3 Ultra nvfp4 1,427
  103. 103 GPT-5 Chat 1,426
  104. 104 Mistral Medium 3.5 1,426
  105. 105 Claude Opus 4 thinking 16k 1,426
  106. 106 DeepSeek-V3.2-Exp thinking 1,425
  107. 107 Muse Glimmer 1,425
  108. 108 Claude Haiku 5.5 high 1,425
  109. 109 DeepSeek-V3.2 1,425
  110. 110 GLM-4.6 1,424
  111. 111 Qwen3-235B-A22B-Instruct Jul 2025 1,422
  112. 112 DeepSeek-R1 May 2025 1,421
  113. 113 Kimi K2 Sep 2025 1,418
  114. 114 Grok 4 Fast non-reasoning 1,418
  115. 115 ERNIE-5.0-Preview (1022) 1,418
  116. 116 Kimi K2 Jul 2025 1,418
  117. 117 DeepSeek-V3.1 1,417
  118. 118 Qwen3.5-122B-A10B 1,416
  119. 119 DeepSeek-V3.1-Terminus thinking 1,416
  120. 120 MiniMax-M2.7 1,415
  121. 121 GPT-4.1 1,415
  122. 122 Claude Haiku 4.5 1,415
  123. 123 Mistral Large 3 1,414
  124. 124 Qwen3-VL-235B-A22B-Instruct 1,413
  125. 125 amazon-nova-experimental-chat (26-01-10) 1,413
  126. 126 Nemotron 3.5 Super VL nvfp4 1,412
  127. 127 Grok 3 1,411
  128. 128 Grok 4 1,411
  129. 129 GLM-4.5 1,411
  130. 130 Gemini 2.5 Flash Jun 2025 1,409
  131. 131 Qwen3.5-27B 1,409
  132. 132 Tencent Hy3 preview 1,409
  133. 133 Mistral Medium 3.1 1,408
  134. 134 Inkling-Small 1,406
  135. 135 Gemini 2.5 Flash Sep 2025 1,403
  136. 136 o1 1,402
  137. 137 Qwen3-235B-A22B non-thinking 1,402
  138. 138 Claude Sonnet 4 thinking 32k 1,402
  139. 139 GPT-5.4 Nano high 1,401
  140. 140 LongCat-Flash-Chat 1,401
  141. 141 Qwen3-235B-A22B-Thinking Jul 2025 1,400
  142. 142 Qwen3-Next-80B-A3B-Instruct 1,399
  143. 143 DeepSeek-R1 1,398
  144. 144 Qwen3.5-Flash 1,396
  145. 145 DeepSeek-V3 Mar 2025 1,395
  146. 146 Qwen3-VL-235B-A22B thinking 1,395
  147. 147 Hunyuan Vision 1.5 Thinking 1,394
  148. 148 Qwen3.5-35B-A3B 1,394
  149. 149 Step 3.5 Flash 1,393
  150. 150 amazon-nova-experimental-chat (12-10) 1,393
  151. 151 MiMo-V2-Flash non-thinking 1,392
  152. 152 o4-mini 1,391
  153. 153 MiniMax-M2.5 1,391
  154. 154 GPT-5 mini high 1,390
  155. 155 o1-preview 1,389
  156. 156 Claude 3.7 Sonnet thinking 32k 1,388
  157. 157 Solar Pro 4 1,388
  158. 158 Qwen3-Coder-480B-A35B 1,388
  159. 159 Hunyuan-T1 1,387
  160. 160 Mistral Medium 3 1,387
  161. 161 MiniMax-M2.1 1,385
  162. 162 Hunyuan-TurboS 1,383
  163. 163 GPT-4.1 mini 1,383
  164. 164 Qwen3-30B-A3B-Instruct Jul 2025 1,382
  165. 165 Gemini 2.5 Flash-Lite Sep 2025, non-thinking, 09-2025 1,379
  166. 166 Trinity Large Preview 1,379
  167. 167 GLM-4.6V 1,378
  168. 168 Gemini 2.5 Flash-Lite Jun 2025 1,375
  169. 169 Claude 3.5 Sonnet (October 2024) 1,374
  170. 170 Qwen2.5-Max 1,374
  171. 171 GLM-4.5-Air 1,373
  172. 172 Qwen3-Next-80B-A3B thinking 1,369
  173. 173 Trinity Large Thinking 1,367
  174. 174 Gemma 3 27B 1,365
  175. 175 glm-4.7-flash 1,365
  176. 176 MiniMax-M1-80k 1,364
  177. 177 amazon-nova-experimental-chat (11-10) 1,364
  178. 178 Grok-3 mini high 1,364
  179. 179 o3-mini high 1,364
  180. 180 Gemini 2.0 Flash Feb 2025 1,360
  181. 181 Nemotron 3 Super 1,360
  182. 182 DeepSeek-V3 1,358
  183. 183 Mistral Small 3.2 1,356
  184. 184 INTELLECT-3 1,356
  185. 185 Cohere Command A 1,354
  186. 186 Gemini 2.0 Flash-Lite 1,354
  187. 187 Nemotron 3.5 Lightning 30B-A3B nvfp4 1,353
  188. 188 GLM-4.5V 1,352
  189. 189 gpt-oss-120b 1,352
  190. 190 Gemini 1.5 Pro Sep 2024 1,351
  191. 191 Step 3 1,349
  192. 192 amazon-nova-experimental-chat (10-20) 1,348
  193. 193 Llama-3.1-Nemotron-Ultra-253B-v1 1,348
  194. 194 Qwen3-32B 1,347
  195. 195 amazon-nova-experimental-chat (10-09) 1,346
  196. 196 GPT-4o May 2024 1,346
  197. 197 Qwen-Plus (0125) 1,346
  198. 198 MiniMax-M2 1,344
  199. 199 Claude 3.5 Sonnet 1,343
  200. 200 Ling-flash-2.0 1,343
  201. 201 Llama-3.3-Nemotron-Super-49B-v1.5 1,343
  202. 202 Mercury 2 1,343
  203. 203 GLM-4-Plus (0111) 1,343
  204. 204 Gemma 3 12B 1,342
  205. 205 Hunyuan Turbo (0110) 1,341
  206. 206 Granite 4.2 30B 1,340
  207. 207 GPT-5 nano high 1,338
  208. 208 o1-mini 1,337
  209. 209 Gemini Advanced (0514) 1,336
  210. 210 Grok-2 Aug 2024 1,336
  211. 211 GPT-4o Aug 2024 1,336
  212. 212 QwQ-32B 1,336
  213. 213 Llama 3.1-405B bf16 1,335
  214. 214 Nova 2 Lite 1,335
  215. 215 Step-2 16K Exp (202412) 1,334
  216. 216 Olmo 3.1 32B Instruct 1,329
  217. 217 Yi-Lightning 1,328
  218. 218 Llama-3.3-Nemotron-Super-49B-v1 1,328
  219. 219 Llama 4 Maverick 1,327
  220. 220 Qwen3-30B-A3B 1,327
  221. 221 Hunyuan Large 20250210 1,326
  222. 222 Claude 3.5 Haiku 1,325
  223. 223 GPT-4 Turbo Apr 2024 1,325
  224. 224 Gemini 1.5 Pro May 2024 1,324
  225. 225 DeepSeek-V2.5-1210 1,323
  226. 226 Ring-flash-2.0 1,322
  227. 227 GPT-4.1 nano 1,322
  228. 228 Claude 3 Opus 1,322
  229. 229 Llama 4 Scout 1,321
  230. 230 Molmo2-8B 1,321
  231. 231 Step-1o Turbo (202506) 1,320
  232. 232 GLM-4-Plus 1,319
  233. 233 Qwen-Max (0919) 1,318
  234. 234 Llama 3.3 70B Instruct 1,318
  235. 235 GPT-4o mini 1,318
  236. 236 Gemma 3n E4B 1,318
  237. 237 gpt-oss-20b 1,317
  238. 238 Qwen2.5-Plus (1127) 1,315
  239. 239 Mistral Large 2 Jul 2024 1,314
  240. 240 Athene-V2-Chat 1,314
  241. 241 Nemotron 3 Nano 30B-A3B bf16 1,314
  242. 242 GPT-4 Turbo (Jan 2024) 1,313
  243. 243 GPT-4 Turbo Nov 2023 1,313
  244. 244 Hunyuan Standard 20250210 1,311
  245. 245 Gemini 1.5 Flash Sep 2024 1,309
  246. 246 Grok-2 mini 1,308
  247. 247 DeepSeek-V2.5 Sep 2024 1,307
  248. 248 Athene-70B 1,307
  249. 249 Olmo 3 32B Think 1,306
  250. 250 Mercury 1,306
  251. 251 Mistral Large 2 Nov 2024 1,306
  252. 252 Magistral Medium 1.0 1,305
  253. 253 Granite 4.1 8B 1,304
  254. 254 Gemma 3 4B 1,303
  255. 255 Qwen2.5-72B 1,303
  256. 256 Mistral Small 3.1 1,303
  257. 257 Llama-3.1-Nemotron-70B-Instruct 1,299
  258. 258 Hunyuan Large Vision 1,294
  259. 259 Llama 3.1-70B 1,293
  260. 260 Amazon Nova Pro 1,290
  261. 261 Granite 4.2 3B 1,290
  262. 262 Granite 4.2 8B 1,290
  263. 263 Gemma 2 27B 1,289
  264. 264 Jamba 1.5 Large 1,289
  265. 265 Reka Core 1,288
  266. 266 GPT-4 Mar 2023 1,288
  267. 267 Granite 4.0 H Small 1,287
  268. 268 Llama-3.1-Nemotron-51B-Instruct 1,287
  269. 269 Gemini 1.5 Flash May 2024 1,287
  270. 270 Olmo 3.1 32B Think 1,286
  271. 271 Tulu 3 (Tülu 3) 70B 1,286
  272. 272 Claude 3 Sonnet 1,281
  273. 273 Gemma 2 9B IT SimPO 1,280
  274. 274 Nemotron-4-340B-Instruct 1,277
  275. 275 Llama 3-70B 1,277
  276. 276 Command R+ 1,276
  277. 277 GPT-4 Jun 2023 1,276
  278. 278 Mistral Small 3 1,274
  279. 279 GLM-4 (0520) 1,273
  280. 280 Reka Flash 1,272
  281. 281 Qwen2.5-Coder-32B-Instruct 1,271
  282. 282 Aya Expanse 32B 1,267
  283. 283 Gemma 2 9B 1,267
  284. 284 DeepSeek-Coder-V2 236B 1,265
  285. 285 Qwen2-72B 1,262
  286. 286 Claude 3 Haiku 1,262
  287. 287 Amazon Nova Lite 1,260
  288. 288 Gemini 1.5 Flash 8B 1,259
  289. 289 Phi-4 1,256
  290. 290 OLMo 2 32B Instruct 1,252
  291. 291 c4ai-command-r-08-2024 1,250
  292. 292 Mistral Large 1,243
  293. 293 Amazon Nova Micro 1,241
  294. 294 Jamba 1.5 Mini 1,240
  295. 295 Ministral 8B 1,238
  296. 296 Gemini 1.0 Pro 1,237
  297. 297 Qwen1.5-110B 1,234
  298. 298 Reka Flash 21B (Online) 1,234
  299. 299 Qwen1.5-72B 1,234
  300. 300 Hunyuan Standard 256K 1,233
  301. 301 Mixtral 8x22B 1,230
  302. 302 Command R (03-2024) 1,227
  303. 303 GPT-3.5 Turbo Jan 2024 1,226
  304. 304 Llama 3-8B 1,224
  305. 305 Aya Expanse 8B 1,223
  306. 306 Mistral Medium 1,223
  307. 307 Llama-3.1-Tulu-3-8B 1,221
  308. 308 Zephyr ORPO 141B A35B v0.1 1,213
  309. 309 Yi-1.5-34B 1,213
  310. 310 Llama 3.1-8B 1,211
  311. 311 Granite-3.1-8B-Instruct 1,209
  312. 312 GPT-3.5 Turbo Nov 2023 1,204
  313. 313 Qwen1.5-32B 1,204
  314. 314 Gemma 2 2B IT 1,200
  315. 315 phi-3-medium 14B 1,198
  316. 316 Mixtral 8x7B 1,197
  317. 317 DBRX 1,196
  318. 318 Qwen1.5-14B 1,191
  319. 319 InternLM2.5-20B-Chat 1,191
  320. 320 DeepSeek LLM 67B 1,185
  321. 321 WizardLM 70B 1,185
  322. 322 Yi-34B 1,184
  323. 323 Granite-3.0-8B-Instruct 1,183
  324. 324 OpenChat 3.5 1,183
  325. 325 OpenChat 3.5 (0106) 1,183
  326. 326 Gemma 1.1 7B IT 1,183
  327. 327 Snowflake Arctic Instruct 1,180
  328. 328 Granite-3.1-2B-Instruct 1,179
  329. 329 Tulu 2 DPO 70B 1,178
  330. 330 OpenHermes 2.5 Mistral 7B 1,176
  331. 331 Vicuna 33B 1,173
  332. 332 phi-3-small 7.4B 1,171
  333. 333 Starling-LM-7B-beta 1,171
  334. 334 Llama 2-70B 1,171
  335. 335 Starling-LM-7B-alpha 1,167
  336. 336 Llama 3.2 3B 1,167
  337. 337 Nous-Hermes-2-Mixtral-8x7B-DPO 1,164
  338. 338 Granite-3.0-2B-Instruct 1,157
  339. 339 Llama2-70B-SteerLM-Chat 1,155
  340. 340 QwQ-32B-Preview 1,154
  341. 341 SOLAR-10.7B-Instruct-v1.0 1,152
  342. 342 Dolphin 2.2.1 Mistral 7B 1,152
  343. 343 MPT-30B 1,151
  344. 344 WizardLM 13B 1,150
  345. 345 Mistral 7B v0.2 1,149
  346. 346 Falcon-180B 1,148
  347. 347 Qwen1.5-7B 1,144
  348. 348 Phi-3 Mini 4K Instruct (June 2024) 1,143
  349. 349 Vicuna 13B 1,142
  350. 350 Llama 2-13B 1,141
  351. 351 Qwen-14B 1,139
  352. 352 PaLM 2 1,139
  353. 353 Gemma 7B 1,138
  354. 354 Code Llama 34B Instruct 1,137
  355. 355 Zephyr 7B Beta 1,131
  356. 356 Phi-3-mini-128k-instruct 1,130
  357. 357 phi-3-mini 3.8B 1,128
  358. 358 Guanaco 33B 1,128
  359. 359 Zephyr 7B Alpha 1,127
  360. 360 StripedHyena Nous 7B 1,122
  361. 361 Code Llama 70B Instruct 1,119
  362. 362 Gemma 1.1 2B IT 1,117
  363. 363 Vicuna 7B 1,115
  364. 364 SmolLM2 1.7B Instruct 1,115
  365. 365 Llama 3.2 1B 1,111
  366. 366 Mistral-7B-Instruct-v0.1 1,110
  367. 367 Llama 2-7B 1,108
  368. 368 Gemma 2B 1,094
  369. 369 Qwen1.5-4B-Chat 1,091
  370. 370 OLMo-7B-Instruct 1,074
  371. 371 Koala 13B 1,071
  372. 372 Alpaca 13B 1,070
  373. 373 GPT4All 13B Snoozy 1,068
  374. 374 MPT-7B 1,064
  375. 375 ChatGLM3-6B 1,057
  376. 376 RWKV-4 Raven 14B 1,042
  377. 377 chatglm2-6b 1,025
  378. 378 OpenAssistant Pythia 12B 1,024
  379. 379 ChatGLM-6B 996
  380. 380 FastChat-T5-3B 993
  381. 381 Dolly 2.0-12b 983
  382. 382 LLaMA-13B 975
  383. 383 stablelm-tuned-alpha-7b 953
#EntryScoreSettingDateSource
1 Gemini 4 Carbon 1,528 high Oct 2026 Official leaderboard
2 Gemini 4 Argon 1,520 high Sep 2026 Official leaderboard
3 Claude Opus 5.5 1,509 high Sep 2026 Official leaderboard
4 Muse Wonder 1,507 max Oct 2026 Official leaderboard
5 Claude Fable 5.1 1,504 max Sep 2026 Official leaderboard
6 Claude Fable 5 1,504 high Jun 2026 Official leaderboard
7 Claude Opus 4.6 1,503 high Feb 2026 Official leaderboard
8 Claude Opus 4.7 1,501 high Apr 2026 Official leaderboard
9 Gemini 3.8 Flash 1,496 high Sep 2026 Official leaderboard
10 Muse Spark 1.3 1,495 max Sep 2026 Official leaderboard
11 Muse Spark 1.2 1,492 xhigh Aug 2026 Official leaderboard
12 Claude Opus 5 1,492 high Jul 2026 Official leaderboard
13 Muse Spark 1.1 1,490 Jul 2026 Official leaderboard
14 Gemini 3.7 Flash 1,490 high Aug 2026 Official leaderboard
15 Muse Spark 1,489 Apr 2026 Official leaderboard
16 Kimi K3 1,488 max Jul 2026 Official leaderboard
17 Gemini 3.1 Pro 1,487 Feb 2026 Official leaderboard
18 Gemini 3 Pro 1,485 Nov 2025 Official leaderboard
19 GPT-5.6 Sol 1,485 xhigh Jul 2026 Official leaderboard
20 Gemini 3.6 Flash 1,483 high Jul 2026 Official leaderboard
21 Qwen 3.8 Max 1,482 Aug 2026 Official leaderboard
22 Claude Opus 4.8 1,482 high May 2026 Official leaderboard
23 GPT-5.5 1,482 high Apr 2026 Official leaderboard
24 GPT-6.1 Sol 1,481 max Sep 2026 Official leaderboard
25 GLM-5.3 1,481 max Aug 2026 Official leaderboard
26 MiMo-V2.6 Pro 1,480 Sep 2026 Official leaderboard
27 Gemini 3.5 Flash 1,478 high May 2026 Official leaderboard
28 Claude Sonnet 5.5 1,477 xhigh Sep 2026 Official leaderboard
29 GLM-5.3-Flash 1,477 Aug 2026 Official leaderboard
30 DeepSeek V4.1 Flash 1,476 max Sep 2026 Official leaderboard
31 GPT-5.2 Instant 1,476 Dec 2025 Official leaderboard
32 GPT-6 Astra 1,476 max Sep 2026 Official leaderboard
33 GLM-5.2 1,475 max Jun 2026 Official leaderboard
34 GPT-5.4 1,475 high Mar 2026 Official leaderboard
35 Grok 4.20 Beta 1 1,475 Oct 2026 Official leaderboard
36 Qwen3.7-Max 1,474 May 2026 Official leaderboard
37 Claude Opus 4.5 1,474 high, 32k Nov 2025 Official leaderboard
38 GPT-5.5 Instant 1,473 May 2026 Official leaderboard
39 Gemini 3 Flash 1,473 Dec 2025 Official leaderboard
40 Claude Sonnet 4.6 1,472 Feb 2026 Official leaderboard
41 Grok 4.20 Beta (Reasoning, 0309) 1,472 Oct 2026 Official leaderboard
42 Grok 4.20 Multi-Agent Beta (0309) 1,471 Oct 2026 Official leaderboard
43 GPT-5.6 Terra 1,467 xhigh Jul 2026 Official leaderboard
44 ERNIE 5.1 1,467 May 2026 Official leaderboard
45 MiMo-V2.5-Pro 1,467 Apr 2026 Official leaderboard
46 Grok 4.5 1,465 Jul 2026 Official leaderboard
47 GLM-5.1 1,465 Apr 2026 Official leaderboard
48 DeepSeek-V4-Pro 1,465 high Apr 2026 Official leaderboard
49 Grok 4.1 1,465 thinking Nov 2025 Official leaderboard
50 Qwen3.5-Max-Preview 1,465 Oct 2026 Official leaderboard
51 Claude Sonnet 5 1,462 high Jun 2026 Official leaderboard
52 Kimi K2.6 1,461 Apr 2026 Official leaderboard
53 Qwen 3.6 Max (Preview) 1,460 Apr 2026 Official leaderboard
54 GPT-6 Sol 1,458 max Sep 2026 Official leaderboard
55 GLM-5 1,457 Feb 2026 Official leaderboard
56 Claude Sonnet 4.5 1,456 high, 32k Sep 2025 Official leaderboard
57 Seed2.0 Pro 1,456 Feb 2026 Official leaderboard
58 GPT-5.1 1,456 high Nov 2025 Official leaderboard
59 Qwen3.7-Plus 1,456 Jun 2026 Official leaderboard
60 Gemini 3.5 Flash-Lite 1,456 Jul 2026 Official leaderboard
61 Grok 4.6 1,455 high Aug 2026 Official leaderboard
62 Tencent Hy3 1,454 Jul 2026 Official leaderboard
63 GPT-5.6 Luna 1,453 xhigh Jul 2026 Official leaderboard
64 Gemma 4 31B IT 1,452 Apr 2026 Official leaderboard
65 GPT-5.3 Instant 1,450 Mar 2026 Official leaderboard
66 Claude Opus 4.1 1,450 thinking 16k Aug 2025 Official leaderboard
67 Kimi K2.5 1,450 thinking Jan 2026 Official leaderboard
68 ERNIE-5.0-Preview (1203) 1,448 Oct 2026 Official leaderboard
69 Step 5 Preview 1,448 Sep 2026 Official leaderboard
70 MiMo-V2-Pro 1,448 Mar 2026 Official leaderboard
71 MiMo-V2.6 Flash 1,448 Sep 2026 Official leaderboard
72 GPT-5.4 Mini 1,447 high Mar 2026 Official leaderboard
73 ERNIE 5.0 (0110) 1,446 Oct 2026 Official leaderboard
74 Gemini 2.5 Pro Jun 2025 1,445 Jun 2025 Official leaderboard
75 GPT-4.5 1,445 Feb 2025 Official leaderboard
76 Qwen 3.6 Plus 1,443 Mar 2026 Official leaderboard
77 GPT-4o Mar 2025 1,443 Mar 2025 Official leaderboard
78 Inkling 1,442 Jul 2026 Official leaderboard
79 Grok 4.3 1,442 Apr 2026 Official leaderboard
80 Qwen3.5 397B-A17B 1,441 Feb 2026 Official leaderboard
81 GLM-4.7 1,441 Dec 2025 Official leaderboard
82 Grok 4.7 1,441 xhigh Sep 2026 Official leaderboard
83 GPT-6 Luna 1,441 max Sep 2026 Official leaderboard
84 MiniMax-M3 1,439 Jun 2026 Official leaderboard
85 DeepSeek-V4-Flash 1,439 high Apr 2026 Official leaderboard
86 Gemma 4 26B A4B 1,438 Apr 2026 Official leaderboard
87 GPT-5.2 1,437 high Dec 2025 Official leaderboard
88 LongCat-Flash-Chat Exp (2602) 1,436 Oct 2026 Official leaderboard
89 Beam 1.0 1,436 Oct 2026 Official leaderboard
90 Qwen 3.8 27B 1,436 Aug 2026 Official leaderboard
91 GPT-5 1,435 high Aug 2025 Official leaderboard
92 Qwen3-Max 1,434 Sep 2025 Official leaderboard
93 GLM-5V-Turbo 1,434 Apr 2026 Official leaderboard
94 MiMo-V2.5 1,434 Apr 2026 Official leaderboard
95 Gemini 3.1 Flash-Lite 1,433 Mar 2026 Official leaderboard
96 Mistral Large 4 1,432 Oct 2026 Official leaderboard
97 o3 1,432 Apr 2025 Official leaderboard
98 MiMo-V2-Omni 1,431 Mar 2026 Official leaderboard
99 Grok 4.1 Fast 1,430 reasoning Nov 2025 Official leaderboard
100 Kimi K2 Thinking Turbo 1,430 Oct 2026 Official leaderboard
101 amazon-nova-experimental-chat (26-02-10) 1,427 Oct 2026 Official leaderboard
102 Nemotron 3 Ultra 1,427 nvfp4 Jun 2026 Official leaderboard
103 GPT-5 Chat 1,426 Aug 2025 Official leaderboard
104 Mistral Medium 3.5 1,426 Apr 2026 Official leaderboard
105 Claude Opus 4 1,426 thinking 16k May 2025 Official leaderboard
106 DeepSeek-V3.2-Exp 1,425 thinking Sep 2025 Official leaderboard
107 Muse Glimmer 1,425 Aug 2026 Official leaderboard
108 Claude Haiku 5.5 1,425 high Oct 2026 Official leaderboard
109 DeepSeek-V3.2 1,425 Dec 2025 Official leaderboard
110 GLM-4.6 1,424 Sep 2025 Official leaderboard
111 Qwen3-235B-A22B-Instruct Jul 2025 1,422 Jul 2025 Official leaderboard
112 DeepSeek-R1 May 2025 1,421 May 2025 Official leaderboard
113 Kimi K2 Sep 2025 1,418 Sep 2025 Official leaderboard
114 Grok 4 Fast 1,418 non-reasoning Sep 2025 Official leaderboard
115 ERNIE-5.0-Preview (1022) 1,418 Oct 2026 Official leaderboard
116 Kimi K2 Jul 2025 1,418 Jul 2025 Official leaderboard
117 DeepSeek-V3.1 1,417 Aug 2025 Official leaderboard
118 Qwen3.5-122B-A10B 1,416 Feb 2026 Official leaderboard
119 DeepSeek-V3.1-Terminus 1,416 thinking Sep 2025 Official leaderboard
120 MiniMax-M2.7 1,415 Mar 2026 Official leaderboard
121 GPT-4.1 1,415 Apr 2025 Official leaderboard
122 Claude Haiku 4.5 1,415 Oct 2025 Official leaderboard
123 Mistral Large 3 1,414 Dec 2025 Official leaderboard
124 Qwen3-VL-235B-A22B-Instruct 1,413 Sep 2025 Official leaderboard
125 amazon-nova-experimental-chat (26-01-10) 1,413 Oct 2026 Official leaderboard
126 Nemotron 3.5 Super VL 1,412 nvfp4 Oct 2026 Official leaderboard
127 Grok 3 1,411 Apr 2025 Official leaderboard
128 Grok 4 1,411 Jul 2025 Official leaderboard
129 GLM-4.5 1,411 Aug 2025 Official leaderboard
130 Gemini 2.5 Flash Jun 2025 1,409 Jun 2025 Official leaderboard
131 Qwen3.5-27B 1,409 Feb 2026 Official leaderboard
132 Tencent Hy3 preview 1,409 Apr 2026 Official leaderboard
133 Mistral Medium 3.1 1,408 Aug 2025 Official leaderboard
134 Inkling-Small 1,406 Jul 2026 Official leaderboard
135 Gemini 2.5 Flash Sep 2025 1,403 Sep 2025 Official leaderboard
136 o1 1,402 Dec 2024 Official leaderboard
137 Qwen3-235B-A22B 1,402 non-thinking Apr 2025 Official leaderboard
138 Claude Sonnet 4 1,402 thinking 32k May 2025 Official leaderboard
139 GPT-5.4 Nano 1,401 high Mar 2026 Official leaderboard
140 LongCat-Flash-Chat 1,401 Aug 2025 Official leaderboard
141 Qwen3-235B-A22B-Thinking Jul 2025 1,400 Jul 2025 Official leaderboard
142 Qwen3-Next-80B-A3B-Instruct 1,399 Sep 2025 Official leaderboard
143 DeepSeek-R1 1,398 Jan 2025 Official leaderboard
144 Qwen3.5-Flash 1,396 Feb 2026 Official leaderboard
145 DeepSeek-V3 Mar 2025 1,395 Mar 2025 Official leaderboard
146 Qwen3-VL-235B-A22B 1,395 thinking Sep 2025 Official leaderboard
147 Hunyuan Vision 1.5 Thinking 1,394 Oct 2026 Official leaderboard
148 Qwen3.5-35B-A3B 1,394 Feb 2026 Official leaderboard
149 Step 3.5 Flash 1,393 Feb 2026 Official leaderboard
150 amazon-nova-experimental-chat (12-10) 1,393 Oct 2026 Official leaderboard
151 MiMo-V2-Flash 1,392 non-thinking Dec 2025 Official leaderboard
152 o4-mini 1,391 Apr 2025 Official leaderboard
153 MiniMax-M2.5 1,391 Feb 2026 Official leaderboard
154 GPT-5 mini 1,390 high Aug 2025 Official leaderboard
155 o1-preview 1,389 Sep 2024 Official leaderboard
156 Claude 3.7 Sonnet 1,388 thinking 32k Feb 2025 Official leaderboard
157 Solar Pro 4 1,388 Aug 2026 Official leaderboard
158 Qwen3-Coder-480B-A35B 1,388 Jul 2025 Official leaderboard
159 Hunyuan-T1 1,387 Mar 2025 Official leaderboard
160 Mistral Medium 3 1,387 May 2025 Official leaderboard
161 MiniMax-M2.1 1,385 Dec 2025 Official leaderboard
162 Hunyuan-TurboS 1,383 Feb 2025 Official leaderboard
163 GPT-4.1 mini 1,383 Apr 2025 Official leaderboard
164 Qwen3-30B-A3B-Instruct Jul 2025 1,382 Jul 2025 Official leaderboard
165 Gemini 2.5 Flash-Lite Sep 2025 1,379 non-thinking, 09-2025 Sep 2025 Official leaderboard
166 Trinity Large Preview 1,379 Jan 2026 Official leaderboard
167 GLM-4.6V 1,378 Dec 2025 Official leaderboard
168 Gemini 2.5 Flash-Lite Jun 2025 1,375 Jun 2025 Official leaderboard
169 Claude 3.5 Sonnet (October 2024) 1,374 Oct 2024 Official leaderboard
170 Qwen2.5-Max 1,374 Jan 2025 Official leaderboard
171 GLM-4.5-Air 1,373 Jul 2025 Official leaderboard
172 Qwen3-Next-80B-A3B 1,369 thinking Sep 2025 Official leaderboard
173 Trinity Large Thinking 1,367 Apr 2026 Official leaderboard
174 Gemma 3 27B 1,365 Mar 2025 Official leaderboard
175 glm-4.7-flash 1,365 Jan 2026 Official leaderboard
176 MiniMax-M1-80k 1,364 Jun 2025 Official leaderboard
177 amazon-nova-experimental-chat (11-10) 1,364 Oct 2026 Official leaderboard
178 Grok-3 mini 1,364 high Apr 2025 Official leaderboard
179 o3-mini 1,364 high Jan 2025 Official leaderboard
180 Gemini 2.0 Flash Feb 2025 1,360 Feb 2025 Official leaderboard
181 Nemotron 3 Super 1,360 Mar 2026 Official leaderboard
182 DeepSeek-V3 1,358 Dec 2024 Official leaderboard
183 Mistral Small 3.2 1,356 Jun 2025 Official leaderboard
184 INTELLECT-3 1,356 Nov 2025 Official leaderboard
185 Cohere Command A 1,354 Mar 2025 Official leaderboard
186 Gemini 2.0 Flash-Lite 1,354 Feb 2025 Official leaderboard
187 Nemotron 3.5 Lightning 30B-A3B 1,353 nvfp4 Aug 2026 Official leaderboard
188 GLM-4.5V 1,352 Aug 2025 Official leaderboard
189 gpt-oss-120b 1,352 Aug 2025 Official leaderboard
190 Gemini 1.5 Pro Sep 2024 1,351 Sep 2024 Official leaderboard
191 Step 3 1,349 Jul 2025 Official leaderboard
192 amazon-nova-experimental-chat (10-20) 1,348 Oct 2026 Official leaderboard
193 Llama-3.1-Nemotron-Ultra-253B-v1 1,348 Apr 2025 Official leaderboard
194 Qwen3-32B 1,347 Apr 2025 Official leaderboard
195 amazon-nova-experimental-chat (10-09) 1,346 Oct 2026 Official leaderboard
196 GPT-4o May 2024 1,346 May 2024 Official leaderboard
197 Qwen-Plus (0125) 1,346 Oct 2026 Official leaderboard
198 MiniMax-M2 1,344 Oct 2025 Official leaderboard
199 Claude 3.5 Sonnet 1,343 Jun 2024 Official leaderboard
200 Ling-flash-2.0 1,343 Sep 2025 Official leaderboard
201 Llama-3.3-Nemotron-Super-49B-v1.5 1,343 Jul 2025 Official leaderboard
202 Mercury 2 1,343 Feb 2026 Official leaderboard
203 GLM-4-Plus (0111) 1,343 Oct 2026 Official leaderboard
204 Gemma 3 12B 1,342 Mar 2025 Official leaderboard
205 Hunyuan Turbo (0110) 1,341 Oct 2026 Official leaderboard
206 Granite 4.2 30B 1,340 Aug 2026 Official leaderboard
207 GPT-5 nano 1,338 high Aug 2025 Official leaderboard
208 o1-mini 1,337 Sep 2024 Official leaderboard
209 Gemini Advanced (0514) 1,336 Oct 2026 Official leaderboard
210 Grok-2 Aug 2024 1,336 Aug 2024 Official leaderboard
211 GPT-4o Aug 2024 1,336 Aug 2024 Official leaderboard
212 QwQ-32B 1,336 Mar 2025 Official leaderboard
213 Llama 3.1-405B 1,335 bf16 Jul 2024 Official leaderboard
214 Nova 2 Lite 1,335 Dec 2025 Official leaderboard
215 Step-2 16K Exp (202412) 1,334 Oct 2026 Official leaderboard
216 Olmo 3.1 32B Instruct 1,329 Dec 2025 Official leaderboard
217 Yi-Lightning 1,328 Dec 2024 Official leaderboard
218 Llama-3.3-Nemotron-Super-49B-v1 1,328 Mar 2025 Official leaderboard
219 Llama 4 Maverick 1,327 Apr 2025 Official leaderboard
220 Qwen3-30B-A3B 1,327 Apr 2025 Official leaderboard
221 Hunyuan Large 20250210 1,326 Feb 2025 Official leaderboard
222 Claude 3.5 Haiku 1,325 Oct 2024 Official leaderboard
223 GPT-4 Turbo Apr 2024 1,325 Apr 2024 Official leaderboard
224 Gemini 1.5 Pro May 2024 1,324 May 2024 Official leaderboard
225 DeepSeek-V2.5-1210 1,323 Dec 2024 Official leaderboard
226 Ring-flash-2.0 1,322 Sep 2025 Official leaderboard
227 GPT-4.1 nano 1,322 Apr 2025 Official leaderboard
228 Claude 3 Opus 1,322 Feb 2024 Official leaderboard
229 Llama 4 Scout 1,321 Apr 2025 Official leaderboard
230 Molmo2-8B 1,321 Dec 2025 Official leaderboard
231 Step-1o Turbo (202506) 1,320 Oct 2026 Official leaderboard
232 GLM-4-Plus 1,319 Oct 2026 Official leaderboard
233 Qwen-Max (0919) 1,318 Oct 2026 Official leaderboard
234 Llama 3.3 70B Instruct 1,318 Dec 2024 Official leaderboard
235 GPT-4o mini 1,318 Jul 2024 Official leaderboard
236 Gemma 3n E4B 1,318 Jun 2025 Official leaderboard
237 gpt-oss-20b 1,317 Aug 2025 Official leaderboard
238 Qwen2.5-Plus (1127) 1,315 Oct 2026 Official leaderboard
239 Mistral Large 2 Jul 2024 1,314 Jul 2024 Official leaderboard
240 Athene-V2-Chat 1,314 Nov 2024 Official leaderboard
241 Nemotron 3 Nano 30B-A3B 1,314 bf16 Dec 2025 Official leaderboard
242 GPT-4 Turbo (Jan 2024) 1,313 Jan 2024 Official leaderboard
243 GPT-4 Turbo Nov 2023 1,313 Nov 2023 Official leaderboard
244 Hunyuan Standard 20250210 1,311 Feb 2025 Official leaderboard
245 Gemini 1.5 Flash Sep 2024 1,309 Aug 2024 Official leaderboard
246 Grok-2 mini 1,308 Aug 2024 Official leaderboard
247 DeepSeek-V2.5 Sep 2024 1,307 Sep 2024 Official leaderboard
248 Athene-70B 1,307 Jul 2024 Official leaderboard
249 Olmo 3 32B Think 1,306 Nov 2025 Official leaderboard
250 Mercury 1,306 Jun 2025 Official leaderboard
251 Mistral Large 2 Nov 2024 1,306 Nov 2024 Official leaderboard
252 Magistral Medium 1.0 1,305 Jun 2025 Official leaderboard
253 Granite 4.1 8B 1,304 Apr 2026 Official leaderboard
254 Gemma 3 4B 1,303 Mar 2025 Official leaderboard
255 Qwen2.5-72B 1,303 Sep 2024 Official leaderboard
256 Mistral Small 3.1 1,303 Mar 2025 Official leaderboard
257 Llama-3.1-Nemotron-70B-Instruct 1,299 Oct 2024 Official leaderboard
258 Hunyuan Large Vision 1,294 Oct 2026 Official leaderboard
259 Llama 3.1-70B 1,293 Jul 2024 Official leaderboard
260 Amazon Nova Pro 1,290 Dec 2024 Official leaderboard
261 Granite 4.2 3B 1,290 Aug 2026 Official leaderboard
262 Granite 4.2 8B 1,290 Aug 2026 Official leaderboard
263 Gemma 2 27B 1,289 Jun 2024 Official leaderboard
264 Jamba 1.5 Large 1,289 Aug 2024 Official leaderboard
265 Reka Core 1,288 Apr 2024 Official leaderboard
266 GPT-4 Mar 2023 1,288 Mar 2023 Official leaderboard
267 Granite 4.0 H Small 1,287 Oct 2025 Official leaderboard
268 Llama-3.1-Nemotron-51B-Instruct 1,287 Sep 2024 Official leaderboard
269 Gemini 1.5 Flash May 2024 1,287 May 2024 Official leaderboard
270 Olmo 3.1 32B Think 1,286 Dec 2025 Official leaderboard
271 Tulu 3 (Tülu 3) 70B 1,286 Nov 2024 Official leaderboard
272 Claude 3 Sonnet 1,281 Feb 2024 Official leaderboard
273 Gemma 2 9B IT SimPO 1,280 Oct 2026 Official leaderboard
274 Nemotron-4-340B-Instruct 1,277 Jun 2024 Official leaderboard
275 Llama 3-70B 1,277 Apr 2024 Official leaderboard
276 Command R+ 1,276 Aug 2024 Official leaderboard
277 GPT-4 Jun 2023 1,276 Jun 2023 Official leaderboard
278 Mistral Small 3 1,274 Jan 2025 Official leaderboard
279 GLM-4 (0520) 1,273 Oct 2026 Official leaderboard
280 Reka Flash 1,272 Feb 2024 Official leaderboard
281 Qwen2.5-Coder-32B-Instruct 1,271 Nov 2024 Official leaderboard
282 Aya Expanse 32B 1,267 Oct 2024 Official leaderboard
283 Gemma 2 9B 1,267 Jun 2024 Official leaderboard
284 DeepSeek-Coder-V2 236B 1,265 Jun 2024 Official leaderboard
285 Qwen2-72B 1,262 Jun 2024 Official leaderboard
286 Claude 3 Haiku 1,262 Mar 2024 Official leaderboard
287 Amazon Nova Lite 1,260 Dec 2024 Official leaderboard
288 Gemini 1.5 Flash 8B 1,259 Oct 2024 Official leaderboard
289 Phi-4 1,256 Dec 2024 Official leaderboard
290 OLMo 2 32B Instruct 1,252 Mar 2025 Official leaderboard
291 c4ai-command-r-08-2024 1,250 Aug 2024 Official leaderboard
292 Mistral Large 1,243 Feb 2024 Official leaderboard
293 Amazon Nova Micro 1,241 Dec 2024 Official leaderboard
294 Jamba 1.5 Mini 1,240 Aug 2024 Official leaderboard
295 Ministral 8B 1,238 Oct 2024 Official leaderboard
296 Gemini 1.0 Pro 1,237 Dec 2023 Official leaderboard
297 Qwen1.5-110B 1,234 Apr 2024 Official leaderboard
298 Reka Flash 21B (Online) 1,234 Feb 2024 Official leaderboard
299 Qwen1.5-72B 1,234 Feb 2024 Official leaderboard
300 Hunyuan Standard 256K 1,233 Oct 2026 Official leaderboard
301 Mixtral 8x22B 1,230 Apr 2024 Official leaderboard
302 Command R (03-2024) 1,227 Mar 2024 Official leaderboard
303 GPT-3.5 Turbo Jan 2024 1,226 Jan 2024 Official leaderboard
304 Llama 3-8B 1,224 Apr 2024 Official leaderboard
305 Aya Expanse 8B 1,223 Oct 2024 Official leaderboard
306 Mistral Medium 1,223 Dec 2023 Official leaderboard
307 Llama-3.1-Tulu-3-8B 1,221 Nov 2024 Official leaderboard
308 Zephyr ORPO 141B A35B v0.1 1,213 Apr 2024 Official leaderboard
309 Yi-1.5-34B 1,213 May 2024 Official leaderboard
310 Llama 3.1-8B 1,211 Jul 2024 Official leaderboard
311 Granite-3.1-8B-Instruct 1,209 Dec 2024 Official leaderboard
312 GPT-3.5 Turbo Nov 2023 1,204 Nov 2023 Official leaderboard
313 Qwen1.5-32B 1,204 Feb 2024 Official leaderboard
314 Gemma 2 2B IT 1,200 Jul 2024 Official leaderboard
315 phi-3-medium 14B 1,198 Apr 2024 Official leaderboard
316 Mixtral 8x7B 1,197 Dec 2023 Official leaderboard
317 DBRX 1,196 Mar 2024 Official leaderboard
318 Qwen1.5-14B 1,191 Feb 2024 Official leaderboard
319 InternLM2.5-20B-Chat 1,191 Aug 2024 Official leaderboard
320 DeepSeek LLM 67B 1,185 Nov 2023 Official leaderboard
321 WizardLM 70B 1,185 Oct 2026 Official leaderboard
322 Yi-34B 1,184 Nov 2023 Official leaderboard
323 Granite-3.0-8B-Instruct 1,183 Oct 2024 Official leaderboard
324 OpenChat 3.5 1,183 Oct 2026 Official leaderboard
325 OpenChat 3.5 (0106) 1,183 Oct 2026 Official leaderboard
326 Gemma 1.1 7B IT 1,183 Apr 2024 Official leaderboard
327 Snowflake Arctic Instruct 1,180 Apr 2024 Official leaderboard
328 Granite-3.1-2B-Instruct 1,179 Dec 2024 Official leaderboard
329 Tulu 2 DPO 70B 1,178 Nov 2023 Official leaderboard
330 OpenHermes 2.5 Mistral 7B 1,176 Oct 2026 Official leaderboard
331 Vicuna 33B 1,173 Oct 2026 Official leaderboard
332 phi-3-small 7.4B 1,171 Apr 2024 Official leaderboard
333 Starling-LM-7B-beta 1,171 Mar 2024 Official leaderboard
334 Llama 2-70B 1,171 Jul 2023 Official leaderboard
335 Starling-LM-7B-alpha 1,167 Nov 2023 Official leaderboard
336 Llama 3.2 3B 1,167 Sep 2024 Official leaderboard
337 Nous-Hermes-2-Mixtral-8x7B-DPO 1,164 Jan 2024 Official leaderboard
338 Granite-3.0-2B-Instruct 1,157 Oct 2024 Official leaderboard
339 Llama2-70B-SteerLM-Chat 1,155 Nov 2023 Official leaderboard
340 QwQ-32B-Preview 1,154 Nov 2024 Official leaderboard
341 SOLAR-10.7B-Instruct-v1.0 1,152 Dec 2023 Official leaderboard
342 Dolphin 2.2.1 Mistral 7B 1,152 Oct 2026 Official leaderboard
343 MPT-30B 1,151 Jun 2023 Official leaderboard
344 WizardLM 13B 1,150 Oct 2026 Official leaderboard
345 Mistral 7B v0.2 1,149 Dec 2023 Official leaderboard
346 Falcon-180B 1,148 Sep 2023 Official leaderboard
347 Qwen1.5-7B 1,144 Feb 2024 Official leaderboard
348 Phi-3 Mini 4K Instruct (June 2024) 1,143 Oct 2026 Official leaderboard
349 Vicuna 13B 1,142 Oct 2026 Official leaderboard
350 Llama 2-13B 1,141 Jul 2023 Official leaderboard
351 Qwen-14B 1,139 Sep 2023 Official leaderboard
352 PaLM 2 1,139 Oct 2026 Official leaderboard
353 Gemma 7B 1,138 Feb 2024 Official leaderboard
354 Code Llama 34B Instruct 1,137 Aug 2023 Official leaderboard
355 Zephyr 7B Beta 1,131 Oct 2023 Official leaderboard
356 Phi-3-mini-128k-instruct 1,130 Apr 2024 Official leaderboard
357 phi-3-mini 3.8B 1,128 Apr 2024 Official leaderboard
358 Guanaco 33B 1,128 Oct 2026 Official leaderboard
359 Zephyr 7B Alpha 1,127 Oct 2023 Official leaderboard
360 StripedHyena Nous 7B 1,122 Oct 2026 Official leaderboard
361 Code Llama 70B Instruct 1,119 Jan 2024 Official leaderboard
362 Gemma 1.1 2B IT 1,117 Apr 2024 Official leaderboard
363 Vicuna 7B 1,115 Oct 2026 Official leaderboard
364 SmolLM2 1.7B Instruct 1,115 Oct 2024 Official leaderboard
365 Llama 3.2 1B 1,111 Sep 2024 Official leaderboard
366 Mistral-7B-Instruct-v0.1 1,110 Sep 2023 Official leaderboard
367 Llama 2-7B 1,108 Jul 2023 Official leaderboard
368 Gemma 2B 1,094 Feb 2024 Official leaderboard
369 Qwen1.5-4B-Chat 1,091 Feb 2024 Official leaderboard
370 OLMo-7B-Instruct 1,074 Feb 2024 Official leaderboard
371 Koala 13B 1,071 Oct 2026 Official leaderboard
372 Alpaca 13B 1,070 Oct 2026 Official leaderboard
373 GPT4All 13B Snoozy 1,068 Oct 2026 Official leaderboard
374 MPT-7B 1,064 May 2023 Official leaderboard
375 ChatGLM3-6B 1,057 Oct 2023 Official leaderboard
376 RWKV-4 Raven 14B 1,042 Oct 2026 Official leaderboard
377 chatglm2-6b 1,025 Jun 2023 Official leaderboard
378 OpenAssistant Pythia 12B 1,024 Oct 2026 Official leaderboard
379 ChatGLM-6B 996 Mar 2023 Official leaderboard
380 FastChat-T5-3B 993 Apr 2023 Official leaderboard
381 Dolly 2.0-12b 983 Apr 2023 Official leaderboard
382 LLaMA-13B 975 Feb 2023 Official leaderboard
383 stablelm-tuned-alpha-7b 953 Apr 2023 Official leaderboard

Labs report scores for 13 models in their cards. We plot every source in the charts below.

More effort adds a median 1 elo

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
Claude Opus 4.6
high
1,503
Claude Opus 4.7
high
1,501
Claude Opus 5
high
1,492
Claude Opus 4.8
high
1,482
GPT-5.5
high
1,482
Gemini 3.5 Flash
medium
1,478
GPT-5.4
high
1,475
Claude Opus 4.5
high, 32k
1,474
Gemini 3 Flash
minimal
1,473
DeepSeek-V4-Pro
high high
1,465
Grok 4.1
thinking
1,465
Claude Sonnet 4.5
high, 32k
1,456
ModelEffortHarnessScoreCost per taskOutput tokens
Gemini 4 Carbon high 1,528
Gemini 4 Argon high 1,520
Claude Opus 5.5 high 1,509
Muse Wonder max 1,507
Claude Fable 5.1 max 1,504
Claude Fable 5 high 1,504
Claude Opus 4.6 high 1,503
Claude Opus 4.7 high 1,501
Claude Opus 4.6 1,496
Gemini 3.8 Flash high 1,496
Muse Spark 1.3 max 1,495
Claude Opus 4.7 1,494
Muse Spark 1.2 xhigh 1,492
Claude Opus 5 high 1,492
Claude Opus 5 max 1,491
Muse Spark 1.1 1,490
Gemini 3.7 Flash high 1,490
Muse Spark 1,489
Kimi K3 max 1,488
Gemini 3.1 Pro 1,487
Gemini 3 Pro 1,485
GPT-5.6 Sol xhigh 1,485
Gemini 3.6 Flash high 1,483
Qwen 3.8 Max 1,482
Claude Opus 4.8 high 1,482
GPT-5.5 high 1,482
GPT-6.1 Sol max 1,481
GLM-5.3 max 1,481
MiMo-V2.6 Pro 1,480
Gemini 3.5 Flash high 1,478
Claude Sonnet 5.5 xhigh 1,477
GPT-5.5 1,477
GLM-5.3-Flash 1,477
Gemini 3.5 Flash medium 1,476
DeepSeek V4.1 Flash max 1,476
GPT-5.2 Instant 1,476
GPT-6 Astra max 1,476
GLM-5.2 max 1,475
GPT-5.4 high 1,475
Claude Opus 4.8 1,474
Qwen3.7-Max 1,474
Claude Opus 4.5 high, 32k 1,474
GPT-5.5 Instant 1,473
Gemini 3 Flash 1,473
Claude Sonnet 4.6 1,472
Claude Opus 4.5 1,470
GPT-5.6 Terra xhigh 1,467
ERNIE 5.1 1,467
MiMo-V2.5-Pro 1,467
Grok 4.5 1,465
GLM-5.1 1,465
DeepSeek-V4-Pro high 1,465
Grok 4.1 thinking 1,465
GPT-5.4 1,465
Claude Sonnet 5 high 1,462
Kimi K2.6 1,461
Qwen 3.6 Max (Preview) 1,460
Grok 4.1 1,459
Gemini 3 Flash minimal 1,458
GPT-6 Sol max 1,458
DeepSeek-V4-Pro 1,458
GLM-5 1,457
Claude Sonnet 4.5 high, 32k 1,456
Seed2.0 Pro 1,456
GPT-5.1 high 1,456
Qwen3.7-Plus 1,456
Gemini 3.5 Flash-Lite 1,456
Grok 4.6 high 1,455
Claude Sonnet 4.5 1,455
DeepSeek-V4-Pro high 1,455
Tencent Hy3 1,454
GPT-5.6 Luna xhigh 1,453
Gemma 4 31B IT 1,452
GPT-5.3 Instant 1,450
Claude Opus 4.1 thinking 16k 1,450
Kimi K2.5 thinking 1,450
Step 5 Preview 1,448
Claude Opus 4.1 1,448
MiMo-V2-Pro 1,448
MiMo-V2.6 Flash 1,448
GPT-5.4 Mini high 1,447
Gemini 2.5 Pro Jun 2025 1,445
GPT-4.5 1,445
Qwen 3.6 Plus 1,443
GPT-4o Mar 2025 1,443
Inkling 1,442
Grok 4.3 1,442
Qwen3.5 397B-A17B 1,441
GLM-4.7 1,441
Grok 4.7 xhigh 1,441
GPT-6 Luna max 1,441
MiniMax-M3 1,439
DeepSeek-V4-Flash high 1,439
GPT-5.1 1,439
Gemma 4 26B A4B 1,438
GPT-5.2 high 1,437
DeepSeek-V4-Flash 1,436
Beam 1.0 1,436
Qwen 3.8 27B 1,436
GPT-5.2 1,436
GPT-5 high 1,435
Qwen3-Max 1,434
GLM-5V-Turbo 1,434
MiMo-V2.5 1,434
Gemini 3.1 Flash-Lite 1,433
Mistral Large 4 1,432
o3 1,432
MiMo-V2-Omni 1,431
Kimi K2.5 instant 1,431
Grok 4.1 Fast reasoning 1,430
Nemotron 3 Ultra nvfp4 1,427
GPT-5 Chat 1,426
Mistral Medium 3.5 1,426
Claude Opus 4 thinking 16k 1,426
DeepSeek-V3.2-Exp thinking 1,425
Muse Glimmer 1,425
Claude Haiku 5.5 high 1,425
DeepSeek-V3.2 1,425
GLM-4.6 1,424
Qwen3-Max 1,423
DeepSeek-V3.2 thinking 1,422
Qwen3-235B-A22B-Instruct Jul 2025 1,422
DeepSeek-V3.2-Exp 1,422
DeepSeek-R1 May 2025 1,421
Kimi K2 Sep 2025 1,418
Grok 4 Fast non-reasoning 1,418
Kimi K2 Jul 2025 1,418
DeepSeek-V3.1 1,417
Qwen3.5-122B-A10B 1,416
DeepSeek-V3.1-Terminus thinking 1,416
DeepSeek-V3.1 thinking 1,416
MiniMax-M2.7 1,415
DeepSeek-V3.1-Terminus 1,415
GPT-4.1 1,415
Claude Haiku 4.5 1,415
Mistral Large 3 1,414
Claude Opus 4 1,414
Qwen3-VL-235B-A22B-Instruct 1,413
Nemotron 3.5 Super VL nvfp4 1,412
Grok 3 1,411
Grok 4 1,411
GLM-4.5 1,411
Gemini 2.5 Flash Jun 2025 1,409
Qwen3.5-27B 1,409
Tencent Hy3 preview 1,409
Mistral Medium 3.1 1,408
Inkling-Small 1,406
Grok 4 Fast reasoning 1,405
Gemini 2.5 Flash Sep 2025 1,403
o1 1,402
Qwen3-235B-A22B non-thinking 1,402
Claude Sonnet 4 thinking 32k 1,402
GPT-5.4 Nano high 1,401
LongCat-Flash-Chat 1,401
Qwen3-235B-A22B-Thinking Jul 2025 1,400
Qwen3-Next-80B-A3B-Instruct 1,399
DeepSeek-R1 1,398
Qwen3.5-Flash 1,396
DeepSeek-V3 Mar 2025 1,395
Qwen3-VL-235B-A22B thinking 1,395
Qwen3.5-35B-A3B 1,394
Step 3.5 Flash 1,393
MiMo-V2-Flash non-thinking 1,392
o4-mini 1,391
Claude Sonnet 4 1,391
MiniMax-M2.5 1,391
GPT-5 mini high 1,390
o1-preview 1,389
Claude 3.7 Sonnet thinking 32k 1,388
Solar Pro 4 1,388
Qwen3-Coder-480B-A35B 1,388
Hunyuan-T1 1,387
Mistral Medium 3 1,387
MiMo-V2-Flash thinking 1,386
MiniMax-M2.1 1,385
Hunyuan-TurboS 1,383
GPT-4.1 mini 1,383
Qwen3-30B-A3B-Instruct Jul 2025 1,382
Gemini 2.5 Flash-Lite Sep 2025 non-thinking, 09-2025 1,379
Trinity Large Preview 1,379
GLM-4.6V 1,378
Qwen3-235B-A22B 1,375
Gemini 2.5 Flash-Lite Jun 2025 1,375
Claude 3.5 Sonnet (October 2024) 1,374
Qwen2.5-Max 1,374
GLM-4.5-Air 1,373
Claude 3.7 Sonnet 1,372
Qwen3-Next-80B-A3B thinking 1,369
Trinity Large Thinking 1,367
Gemma 3 27B 1,365
glm-4.7-flash 1,365
MiniMax-M1-80k 1,364
Grok-3 mini high 1,364
o3-mini high 1,364
Gemini 2.0 Flash Feb 2025 1,360
Nemotron 3 Super 1,360
DeepSeek-V3 1,358
Grok-3 mini 1,358
Mistral Small 3.2 1,356
INTELLECT-3 1,356
Cohere Command A 1,354
Gemini 2.0 Flash-Lite 1,354
Nemotron 3.5 Lightning 30B-A3B nvfp4 1,353
GLM-4.5V 1,352
gpt-oss-120b 1,352
Gemini 1.5 Pro Sep 2024 1,351
Step 3 1,349
Hunyuan-TurboS 1,349
o3-mini 1,348
Llama-3.1-Nemotron-Ultra-253B-v1 1,348
Qwen3-32B 1,347
GPT-4o May 2024 1,346
MiniMax-M2 1,344
Claude 3.5 Sonnet 1,343
Ling-flash-2.0 1,343
Llama-3.3-Nemotron-Super-49B-v1.5 1,343
Mercury 2 1,343
Gemma 3 12B 1,342
Granite 4.2 30B 1,340
GPT-5 nano high 1,338
o1-mini 1,337
Grok-2 Aug 2024 1,336
GPT-4o Aug 2024 1,336
QwQ-32B 1,336
Llama 3.1-405B bf16 1,335
Nova 2 Lite 1,335
Llama 3.1-405B fp8 1,334
Olmo 3.1 32B Instruct 1,329
Yi-Lightning 1,328
Llama-3.3-Nemotron-Super-49B-v1 1,328
Llama 4 Maverick 1,327
Qwen3-30B-A3B 1,327
Claude 3.5 Haiku 1,325
GPT-4 Turbo Apr 2024 1,325
Gemini 1.5 Pro May 2024 1,324
DeepSeek-V2.5-1210 1,323
Ring-flash-2.0 1,322
GPT-4.1 nano 1,322
Claude 3 Opus 1,322
Llama 4 Scout 1,321
Molmo2-8B 1,321
Llama 3.3 70B Instruct 1,318
GPT-4o mini 1,318
Gemma 3n E4B 1,318
gpt-oss-20b 1,317
Mistral Large 2 Jul 2024 1,314
Athene-V2-Chat 1,314
Nemotron 3 Nano 30B-A3B bf16 1,314
GPT-4 Turbo (Jan 2024) 1,313
GPT-4 Turbo Nov 2023 1,313
Gemini 1.5 Flash Sep 2024 1,309
Grok-2 mini 1,308
DeepSeek-V2.5 Sep 2024 1,307
Athene-70B 1,307
Olmo 3 32B Think 1,306
Mercury 1,306
Mistral Large 2 Nov 2024 1,306
Magistral Medium 1.0 1,305
Granite 4.1 8B 1,304
Gemma 3 4B 1,303
Qwen2.5-72B 1,303
Mistral Small 3.1 1,303
Llama-3.1-Nemotron-70B-Instruct 1,299
Llama 3.1-70B 1,293
Amazon Nova Pro 1,290
Granite 4.2 3B 1,290
Granite 4.2 8B 1,290
Gemma 2 27B 1,289
Jamba 1.5 Large 1,289
Reka Core 1,288
GPT-4 Mar 2023 1,288
Granite 4.0 H Small 1,287
Llama-3.1-Nemotron-51B-Instruct 1,287
Gemini 1.5 Flash May 2024 1,287
Olmo 3.1 32B Think 1,286
Tulu 3 (Tülu 3) 70B 1,286
Claude 3 Sonnet 1,281
Nemotron-4-340B-Instruct 1,277
Llama 3-70B 1,277
Command R+ 1,276
GPT-4 Jun 2023 1,276
Mistral Small 3 1,274
Reka Flash 1,272
Qwen2.5-Coder-32B-Instruct 1,271
Aya Expanse 32B 1,267
Gemma 2 9B 1,267
DeepSeek-Coder-V2 236B 1,265
Qwen2-72B 1,262
Command R+ 1,262
Claude 3 Haiku 1,262
Amazon Nova Lite 1,260
Gemini 1.5 Flash 8B 1,259
Phi-4 1,256
OLMo 2 32B Instruct 1,252
c4ai-command-r-08-2024 1,250
Mistral Large 1,243
Amazon Nova Micro 1,241
Jamba 1.5 Mini 1,240
Ministral 8B 1,238
Gemini 1.0 Pro 1,237
Qwen1.5-110B 1,234
Qwen1.5-72B 1,234
Mixtral 8x22B 1,230
Command R (03-2024) 1,227
Reka Flash 1,227
GPT-3.5 Turbo Jan 2024 1,226
Llama 3-8B 1,224
Gemini 1.0 Pro 1,224
Aya Expanse 8B 1,223
Mistral Medium 1,223
Llama-3.1-Tulu-3-8B 1,221
Zephyr ORPO 141B A35B v0.1 1,213
Yi-1.5-34B 1,213
Llama 3.1-8B 1,211
Granite-3.1-8B-Instruct 1,209
GPT-3.5 Turbo Nov 2023 1,204
Qwen1.5-32B 1,204
Gemma 2 2B IT 1,200
phi-3-medium 14B 1,198
Mixtral 8x7B 1,197
DBRX 1,196
Qwen1.5-14B 1,191
InternLM2.5-20B-Chat 1,191
DeepSeek LLM 67B 1,185
Yi-34B 1,184
Granite-3.0-8B-Instruct 1,183
Gemma 1.1 7B IT 1,183
Snowflake Arctic Instruct 1,180
Granite-3.1-2B-Instruct 1,179
Tulu 2 DPO 70B 1,178
phi-3-small 7.4B 1,171
Starling-LM-7B-beta 1,171
Llama 2-70B 1,171
Starling-LM-7B-alpha 1,167
Llama 3.2 3B 1,167
Nous-Hermes-2-Mixtral-8x7B-DPO 1,164
Granite-3.0-2B-Instruct 1,157
Llama2-70B-SteerLM-Chat 1,155
QwQ-32B-Preview 1,154
SOLAR-10.7B-Instruct-v1.0 1,152
MPT-30B 1,151
Mistral 7B v0.2 1,149
Falcon-180B 1,148
Qwen1.5-7B 1,144
Llama 2-13B 1,141
Qwen-14B 1,139
Gemma 7B 1,138
Code Llama 34B Instruct 1,137
Zephyr 7B Beta 1,131
Phi-3-mini-128k-instruct 1,130
phi-3-mini 3.8B 1,128
Zephyr 7B Alpha 1,127
Code Llama 70B Instruct 1,119
Gemma 1.1 2B IT 1,117
SmolLM2 1.7B Instruct 1,115
Llama 3.2 1B 1,111
Mistral-7B-Instruct-v0.1 1,110
Llama 2-7B 1,108
Gemma 2B 1,094
Qwen1.5-4B-Chat 1,091
OLMo-7B-Instruct 1,074
MPT-7B 1,064
ChatGLM3-6B 1,057
chatglm2-6b 1,025
ChatGLM-6B 996
FastChat-T5-3B 993
Dolly 2.0-12b 983
LLaMA-13B 975
stablelm-tuned-alpha-7b 953

20 models have held number one since its release in May 2023

Lab claims count from the day the card came out, independent runs from the model’s release. We draw one row each for the benchmark’s own leaderboard and what labs claim in their cards.

OpenAI Anthropic Google DeepMind xAI Mistral AI Moonshot AI
Own leaderboard 20 leaders
Gemini 4 Carbon 1,528
Lab claims 6 leaders
Kimi K3 1,486

Scores over time

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

On the benchmark’s own leaderboard Reported by the lab in its model card
800 1,000 1,200 1,400 1,600 2023 2024 2025 2026 2027 800 1,000 1,200 1,400 1,600 2023 2024 2025 2026 2027

Labs headline it less often now, last in Dec 2025

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

xAI
Qwen
Mistral AI
DateLaunch postScoreAs the post writes it
2 Dec 2025 Mistral Large 3 1418 LMArena ELO Score (+/- 11)
17 Nov 2025 Grok 4.1 1483 LMArena Text Leaderboard overall Elo (Grok 4.1 Thinking, 'quasarflux')
24 Sep 2025 Qwen3-Max LMArena Text Arena
19 Feb 2025 Grok 3 1402 Chatbot Arena Score (early Grok 3 'chocolate')
13 Aug 2024 Grok-2 LMSYS leaderboard overall Elo (early Grok-2 as 'sus-column-r')

13 system cards report it, from 5 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
15 Oct 2026 Nemotron 3.5 Super VL technical report 1420 Text Arena, Overall, With Style Control, 1420 +/- 12, 2,328 votes, preliminary NVFP4 (Table 11) Capabilities
27 Jul 2026 Kimi K3 technical report 1486 Text Arena (Elo), rank 8 of 200 (Table 5) Other
24 Jul 2026 Gemma 4 31B IT technical report 1451 Arena Text Elo, as of June 19, 2026 (Table 4) Capabilities
24 Jul 2026 Gemma 4 26B A4B technical report 1438 Arena Text Elo, as of June 19, 2026 (Table 4) Capabilities
17 Feb 2026 GLM-5 technical report 1452 LMArena Text Arena score, rank 11 (Figure 3 screenshot); LMsys Chatbot Arena 1452 (Figure 11) Capabilities
21 Jul 2025 Kimi K2 technical report LMSYS Arena text leaderboard as of 2025-07-17: top open-source model and 5th overall, no rating printed (Appendix C) Capabilities
7 Jul 2025 Gemini 2.5 Pro technical report LMArena score, plotted in Figure 1 without printed value; text says +122 points over Gemini 1.5 Pro Capabilities
7 Jul 2025 Gemini 2.5 Flash technical report LMArena score, plotted in Figure 1 without printed value; text says +111 points over Gemini 1.5 Flash Capabilities
25 Mar 2025 Gemma 3 27B technical report 1338 Chatbot Arena Elo, IT, preliminary results received March 8, 2025 (Table 5) Capabilities
31 Jul 2024 Gemma 2 9B technical report 1187 LMSYS Chatbot Arena Elo, IT (Table 14) Capabilities
31 Jul 2024 Gemma 2 27B technical report 1218 LMSYS Chatbot Arena Elo, IT (Table 14) Capabilities
8 Jan 2024 Mixtral 8x7B technical report 1121 LMSys Arena Elo, Mixtral 8x7B Instruct v0.1, screenshot Dec 22, 2023 (Figure 6 caption) Capabilities
10 Oct 2023 Mistral 7B v0.1 technical report 1031 Chatbot Arena ELO Rating, Mistral 7B Instruct (Table 3) Capabilities

More in Instructions and chat

All 76 in Instructions and chat
  • IFEval In use Models answer about 500 prompts containing verifiable instructions such as word limits or required keywords. Code checks each instruction for strict and loose accuracy. Gemma 4 31B IT 98.9%
  • IFBench In use 58 new out-of-domain verifiable output constraints. Scored by programmatic constraint checks (strict or loose). Nemotron-Cascade 2 30B-A3B 82.9%
  • MultiChallenge In use Multi-turn conversations testing instruction retention, inference memory, versioned editing and self-coherence. Graded by LLM judge with rubrics. GPT-6.1 Sol 82%
  • Multi-IF In use Models follow verifiable instructions over three-turn conversations in eight languages, extending IFEval. Scored by instruction accuracy at each turn. LFM3-1.7B 89.1%
  • AdvancedIF In use Models respond to over 1,600 prompts with complex, multi-turn, and system-level instructions. A judge model checks each response against expert-written rubrics. MAI-Thinking-1 85%
  • IFStruct In use Models generate JSON or YAML that must follow a requested schema and phrasing constraints, without constrained decoding. A response passes only if every structural constraint holds. LFM3-1.7B 93.7%