Leaderboards
Benchmarks
Every score carries its origin.
Arena Creative Writing168Arena Instruction Following168Arena Text (overall)168IFEval16LiveBench Instruction Following58LiveBench Language58
Arena Text (overall)
Arena Score (Bradley-Terry, style-controlled) from human pairwise votes — not a percentageProvenance68 of 168 modelsscores Sep 13, 2026 – Sep 25, 2026
101Qwen3.5-35B-A3BQwen36BScore1394Score / $1859Score / GB VRAM~61ScoredSep 25, 2026102Step 3.5 FlashStepFun199BScore1393Score / $4644Score / GB VRAM~11ScoredSep 25, 2026103o4 MiniOpenAIScore1391Score / $316ScoredSep 25, 2026104MiniMax M2.5MiniMax229BScore1391Score / $1464Score / GB VRAM~9.6ScoredSep 25, 2026105Claude Sonnet 4AnthropicScore1391Score / $93ScoredSep 25, 2026106GPT-5 MiniOpenAIScore1390Score / $1390ScoredSep 25, 2026107Qwen3 Coder 480B A35BQwen480BScore1387Score / $1387Score / GB VRAM~4.6ScoredSep 25, 2026108GPT-4.1 MiniOpenAIScore1383Score / $864ScoredSep 25, 2026109Qwen3 30B A3B Instruct 2507Qwen30.5BScore1383Score / $7162Score / GB VRAM~72ScoredSep 25, 2026110GLM 4.6VZ.ai108BScore1379Score / $1532Score / GB VRAM~20ScoredSep 25, 2026111Claude Sonnet 5AnthropicScore1374Score / $137ScoredSep 13, 2026112GLM 4.5 AirZ.ai111BScore1373Score / $1616Score / GB VRAM~20ScoredSep 25, 2026113Qwen3 Next 80B A3B ThinkingQwen81.3BScore1369Score / $1141Score / GB VRAM~27ScoredSep 25, 2026114Trinity Large ThinkingArcee AI399BScore1368Score / $1710Score / GB VRAM~5.4ScoredSep 25, 2026115Gemma 3 27BGoogle27.4BScore1365Score / $8534Score / GB VRAM~79ScoredSep 25, 2026116GLM 4.7 FlashZ.ai31.2BScore1365Score / $3413Score / GB VRAM72ScoredSep 25, 2026117MiniMax M1MiniMax456BScore1364Score / $620Score / GB VRAM~4.7ScoredSep 25, 2026118o3 Mini HighOpenAIScore1364Score / $310ScoredSep 25, 2026119Nemotron 3 SuperNVIDIA124BScore1360Score / $3401Score / GB VRAM~17ScoredSep 25, 2026120DeepSeek V3DeepSeek685BScore1358Score / $1526Score / GB VRAM3ScoredSep 25, 2026121GLM 4.5VZ.ai108BScore1352Score / $751Score / GB VRAM~20ScoredSep 25, 2026122gpt-oss-120bOpenAI120BScore1352Score / $7951Score / GB VRAM~18ScoredSep 25, 2026123o3 MiniOpenAIScore1348Score / $306ScoredSep 25, 2026124Qwen3 32BQwen32.8BScore1347Score / $4811Score / GB VRAM~65ScoredSep 25, 2026125Nemotron 3.5 LightningNVIDIA31.6BScore1347Score / $8417Score / GB VRAM~68ScoredSep 25, 2026126GPT-4o (2024-05-13)OpenAIScore1346Score / $90ScoredSep 25, 2026127Ling-2.6-flashinclusionAI107BScore1344Score / $44800Score / GB VRAM~20ScoredSep 13, 2026128MiniMax M2MiniMax229BScore1343Score / $1317Score / GB VRAM~9.3ScoredSep 25, 2026129Mercury 2InceptionScore1343Score / $1791ScoredSep 25, 2026130Gemma 3 12BGoogle12.2BScore1342Score / $13418Score / GB VRAM~170ScoredSep 25, 2026131GPT-5 NanoOpenAIScore1338Score / $6688ScoredSep 25, 2026132GPT-4o (2024-08-06)OpenAIScore1336Score / $134ScoredSep 25, 2026133Nova 2 LiteAmazonScore1335Score / $534ScoredSep 25, 2026134Llama 4 MaverickMeta402BScore1327Score / $2034Score / GB VRAM~5.2ScoredSep 25, 2026135GPT-4 TurboOpenAIScore1324Score / $44ScoredSep 25, 2026136GPT-4.1 NanoOpenAIScore1322Score / $3306ScoredSep 25, 2026137Llama 4 ScoutMeta109BScore1322Score / $4405Score / GB VRAM~19ScoredSep 25, 2026138Llama 3.3 70B InstructMeta70.6BScore1318Score / $4118Score / GB VRAM~30ScoredSep 25, 2026139GPT-4o-mini (2024-07-18)OpenAIScore1318Score / $2196ScoredSep 25, 2026140gpt-oss-safeguard-20bOpenAI21.5BScore1318Score / $14640Score / GB VRAM~97ScoredSep 25, 2026141Gemma 3n 4BGoogle7.8BScore1317Score / $10978Score / GB VRAM~270ScoredSep 25, 2026142Mistral Large 2407Mistral AIScore1314Score / $219ScoredSep 25, 2026143Nemotron 3 Nano 30B A3BNVIDIA31.6BScore1313Score / $6567Score / GB VRAM~66ScoredSep 25, 2026144Olmo 3 32B ThinkAllen Institute for AI32.2BScore1306Score / $2613Score / GB VRAM~64ScoredSep 25, 2026145Granite 4.1 8BIBM8.8BScore1304Score / $13041Score / GB VRAM~230ScoredSep 25, 2026146Gemma 3 4BGoogle4.3BScore1303Score / $13033Score / GB VRAM~480ScoredSep 25, 2026147Mistral Small 3.1 24BMistral AI24BScore1303Score / $2348Score / GB VRAM~86ScoredSep 25, 2026148Qwen2.5 72B InstructQwen72.7BScore1303Score / $3257Score / GB VRAM~28ScoredSep 25, 2026149Llama 3.1 70B InstructMeta70.6BScore1293Score / $3233Score / GB VRAM~29ScoredSep 25, 2026150Nova Pro 1.0AmazonScore1290Score / $403ScoredSep 25, 2026151Gemma 2 27BGoogle27.2BScore1289Score / $1984Score / GB VRAM~75ScoredSep 25, 2026152Jamba Large 1.7AI21399BScore1289Score / $161Score / GB VRAM~5.1ScoredSep 13, 2026153Granite 4.2 8BIBM8.8BScore1289Score / $8594Score / GB VRAM~230ScoredSep 25, 2026154GPT-4OpenAIScore1288Score / $21ScoredSep 25, 2026155Mistral Small 3Mistral AI23.6BScore1274Score / $15930Score / GB VRAM~86ScoredSep 25, 2026156Qwen2.5 Coder 32B InstructQwen32.8BScore1271Score / $1271Score / GB VRAM~61ScoredSep 25, 2026157Claude 3 HaikuAnthropicScore1262Score / $1009ScoredSep 25, 2026158Nova Lite 1.0AmazonScore1260Score / $5252ScoredSep 25, 2026159Phi 4Microsoft14.7BScore1256Score / $8974Score / GB VRAM139ScoredSep 25, 2026160Command R (08-2024)CohereScore1250Score / $2084ScoredSep 25, 2026161Mistral LargeMistral AIScore1243Score / $207ScoredSep 25, 2026162Nova Micro 1.0AmazonScore1241Score / $8862ScoredSep 25, 2026163Ministral 8BMistral AIScore1238Score / $11251ScoredSep 25, 2026164Mixtral 8x22B InstructMistral AI141BScore1230Score / $205Score / GB VRAM~14ScoredSep 25, 2026165GPT-3.5 Turbo (older v0613)OpenAIScore1226Score / $613ScoredSep 25, 2026166Llama 3.1 8B InstructMeta8BScore1211Score / $30281Score / GB VRAM~240ScoredSep 25, 2026167Llama 3.2 3B InstructMeta3.2BScore1167Score / $23335Score / GB VRAM~580ScoredSep 25, 2026168Llama 3.2 1B InstructMeta1.2BScore1111Score / $55555Score / GB VRAM~1500ScoredSep 25, 2026
Showing 101–168 of 168 models on this board.