Leaderboards
Benchmarks
Every score carries its origin.
Arena Hard Prompts168GPQA Diamond16LiveBench58LiveBench Data Analysis58LiveBench Reasoning58MMLU-Pro16
Arena Hard Prompts
Arena Score on the hard-prompt subset — not a percentageProvenance68 of 168 modelsscores Sep 13, 2026 – Sep 25, 2026
101Qwen3.5-FlashQwenScore1416Score / $5445ScoredSep 25, 2026102Qwen3 Coder 480B A35BQwen480BScore1414Score / $1414Score / GB VRAM~4.7ScoredSep 25, 2026103Step 3.5 FlashStepFun199BScore1413Score / $4709Score / GB VRAM~11ScoredSep 25, 2026104Qwen3.5-35B-A3BQwen36BScore1412Score / $1883Score / GB VRAM~62ScoredSep 25, 2026105DeepSeek V3 0324DeepSeek685BScore1408Score / $1565Score / GB VRAM~3.3ScoredSep 25, 2026106Qwen3 30B A3B Instruct 2507Qwen30.5BScore1407Score / $7290Score / GB VRAM~73ScoredSep 25, 2026107o4 MiniOpenAIScore1406Score / $320ScoredSep 25, 2026108GPT-5 MiniOpenAIScore1404Score / $1404ScoredSep 25, 2026109GPT-4.1 MiniOpenAIScore1402Score / $876ScoredSep 25, 2026110o3 Mini HighOpenAIScore1402Score / $319ScoredSep 25, 2026111Claude Sonnet 5AnthropicScore1399Score / $140ScoredSep 13, 2026112GLM 4.5 AirZ.ai111BScore1391Score / $1636Score / GB VRAM~20ScoredSep 25, 2026113Qwen3 Next 80B A3B ThinkingQwen81.3BScore1386Score / $1155Score / GB VRAM~27ScoredSep 25, 2026114Trinity Large ThinkingArcee AI399BScore1385Score / $1731Score / GB VRAM~5.5ScoredSep 25, 2026115GLM 4.6VZ.ai108BScore1385Score / $1538Score / GB VRAM~20ScoredSep 25, 2026116GLM 4.7 FlashZ.ai31.2BScore1384Score / $3461Score / GB VRAM73ScoredSep 25, 2026117MiniMax M1MiniMax456BScore1381Score / $628Score / GB VRAM~4.8ScoredSep 25, 2026118Nemotron 3 SuperNVIDIA124BScore1378Score / $3445Score / GB VRAM~18ScoredSep 25, 2026119GLM 4.5VZ.ai108BScore1374Score / $763Score / GB VRAM~20ScoredSep 25, 2026120o3 MiniOpenAIScore1371Score / $312ScoredSep 25, 2026121Nemotron 3.5 LightningNVIDIA31.6BScore1369Score / $8556Score / GB VRAM~69ScoredSep 25, 2026122Qwen3 32BQwen32.8BScore1367Score / $4883Score / GB VRAM~66ScoredSep 25, 2026123MiniMax M2MiniMax229BScore1367Score / $1340Score / GB VRAM~9.5ScoredSep 25, 2026124Gemma 3 27BGoogle27.4BScore1366Score / $8540Score / GB VRAM~79ScoredSep 25, 2026125Ling-2.6-flashinclusionAI107BScore1365Score / $45501Score / GB VRAM~20ScoredSep 13, 2026126gpt-oss-120bOpenAI120BScore1362Score / $8011Score / GB VRAM~18ScoredSep 25, 2026127Mercury 2InceptionScore1358Score / $1811ScoredSep 25, 2026128Nova 2 LiteAmazonScore1357Score / $543ScoredSep 25, 2026129GPT-5 NanoOpenAIScore1356Score / $6781ScoredSep 25, 2026130DeepSeek V3DeepSeek685BScore1350Score / $1517Score / GB VRAM3ScoredSep 25, 2026131Llama 4 MaverickMeta402BScore1339Score / $2052Score / GB VRAM~5.3ScoredSep 25, 2026132GPT-4o (2024-05-13)OpenAIScore1338Score / $89ScoredSep 25, 2026133GPT-4.1 NanoOpenAIScore1333Score / $3332ScoredSep 25, 2026134Gemma 3 12BGoogle12.2BScore1332Score / $13320Score / GB VRAM~170ScoredSep 25, 2026135Llama 4 ScoutMeta109BScore1329Score / $4431Score / GB VRAM~19ScoredSep 25, 2026136GPT-4o (2024-08-06)OpenAIScore1327Score / $133ScoredSep 25, 2026137Olmo 3 32B ThinkAllen Institute for AI32.2BScore1326Score / $2652Score / GB VRAM~65ScoredSep 25, 2026138Nemotron 3 Nano 30B A3BNVIDIA31.6BScore1325Score / $6627Score / GB VRAM~67ScoredSep 25, 2026139gpt-oss-safeguard-20bOpenAI21.5BScore1323Score / $14700Score / GB VRAM~98ScoredSep 25, 2026140Llama 3.3 70B InstructMeta70.6BScore1321Score / $4127Score / GB VRAM~30ScoredSep 25, 2026141Granite 4.1 8BIBM8.8BScore1320Score / $13201Score / GB VRAM~240ScoredSep 25, 2026142Mistral Large 2407Mistral AIScore1320Score / $220ScoredSep 25, 2026143Mistral Small 3.1 24BMistral AI24BScore1319Score / $2377Score / GB VRAM~87ScoredSep 25, 2026144Qwen2.5 72B InstructQwen72.7BScore1317Score / $3294Score / GB VRAM~29ScoredSep 25, 2026145GPT-4 TurboOpenAIScore1316Score / $44ScoredSep 25, 2026146Gemma 3n 4BGoogle7.8BScore1312Score / $10936Score / GB VRAM~270ScoredSep 25, 2026147GPT-4o-mini (2024-07-18)OpenAIScore1311Score / $2185ScoredSep 25, 2026148GPT-4OpenAIScore1306Score / $22ScoredSep 25, 2026149Granite 4.2 8BIBM8.8BScore1305Score / $8698Score / GB VRAM~240ScoredSep 25, 2026150Qwen2.5 Coder 32B InstructQwen32.8BScore1303Score / $1303Score / GB VRAM~63ScoredSep 25, 2026151Nova Pro 1.0AmazonScore1303Score / $407ScoredSep 25, 2026152Llama 3.1 70B InstructMeta70.6BScore1299Score / $3247Score / GB VRAM~29ScoredSep 25, 2026153Mistral Small 3Mistral AI23.6BScore1285Score / $16062Score / GB VRAM~86ScoredSep 25, 2026154Gemma 3 4BGoogle4.3BScore1284Score / $12841Score / GB VRAM~470ScoredSep 25, 2026155Jamba Large 1.7AI21399BScore1284Score / $160Score / GB VRAM~5.1ScoredSep 13, 2026156Gemma 2 27BGoogle27.2BScore1282Score / $1973Score / GB VRAM~75ScoredSep 25, 2026157Phi 4Microsoft14.7BScore1278Score / $9128Score / GB VRAM141ScoredSep 25, 2026158Nova Lite 1.0AmazonScore1272Score / $5298ScoredSep 25, 2026159Claude 3 HaikuAnthropicScore1266Score / $1013ScoredSep 25, 2026160Mistral LargeMistral AIScore1262Score / $210ScoredSep 25, 2026161Command R (08-2024)CohereScore1257Score / $2095ScoredSep 25, 2026162Ministral 8BMistral AIScore1251Score / $11372ScoredSep 25, 2026163Nova Micro 1.0AmazonScore1245Score / $8891ScoredSep 25, 2026164Mixtral 8x22B InstructMistral AI141BScore1244Score / $207Score / GB VRAM~14ScoredSep 25, 2026165GPT-3.5 Turbo (older v0613)OpenAIScore1229Score / $615ScoredSep 25, 2026166Llama 3.1 8B InstructMeta8BScore1222Score / $30560Score / GB VRAM~240ScoredSep 25, 2026167Llama 3.2 3B InstructMeta3.2BScore1168Score / $23367Score / GB VRAM~580ScoredSep 25, 2026168Llama 3.2 1B InstructMeta1.2BScore1114Score / $55718Score / GB VRAM~1500ScoredSep 25, 2026
Showing 101–168 of 168 models on this board.