Leaderboards
Benchmarks
Every score carries its origin.
Arena Creative Writing168Arena Instruction Following168Arena Text (overall)168IFEval16LiveBench Instruction Following58LiveBench Language58
Arena Instruction Following
Arena Score on instruction-following prompts — not a percentageProvenance100 of 168 modelsscores Aug 10, 2026 – Sep 25, 2026
1Claude Opus 5.5AnthropicScore1516Score / $76ScoredSep 25, 20262Claude Fable 5AnthropicScore1511Score / $30ScoredSep 13, 20263Claude Opus 4.6AnthropicScore1500Score / $60ScoredSep 25, 20264Claude Fable 5.1AnthropicScore1497Score / $30ScoredSep 25, 20265Claude Opus 5AnthropicScore1497Score / $60ScoredSep 25, 20266Claude Opus 4.7AnthropicScore1494Score / $60ScoredSep 25, 20267GPT-5.6 SolOpenAIScore1489Score / $298ScoredSep 25, 20268Gemini 3.8 FlashGoogleScore1488Score / $794ScoredSep 25, 20269Gemini 3.7 FlashGoogleScore1487Score / $793ScoredSep 25, 202610Kimi K3Moonshot AI2.8TScore1486Score / $151Score / GB VRAM~0.85ScoredSep 25, 202611Muse Spark 1.3MetaScore1486Score / $350ScoredSep 25, 202612GLM 5.3Z.aiScore1480Score / $1298ScoredSep 25, 202613Gemini 3.1 Pro Preview Custom ToolsGoogleScore1479Score / $247ScoredSep 25, 202614DeepSeek V4.1 FlashDeepSeek763BScore1478Score / $3695Score / GB VRAM~3.1ScoredSep 25, 202615Claude Opus 4.8AnthropicScore1478Score / $59ScoredSep 25, 202616Gemini 3.6 FlashGoogleScore1477Score / $788ScoredAug 10, 202617Muse Spark 1.2MetaScore1475Score / $347ScoredSep 25, 202618GPT-5.5OpenAIScore1475Score / $98ScoredSep 25, 202619Claude Sonnet 4.6AnthropicScore1475Score / $98ScoredSep 25, 202620Claude Opus 4.5AnthropicScore1475Score / $59ScoredSep 25, 202621GLM 5.3 FlashZ.ai321BScore1474Score / $5898Score / GB VRAM~7.3ScoredSep 25, 202622GPT-6 AstraOpenAIScore1473Score / $59ScoredSep 25, 202623Muse Spark 1.1MetaScore1473Score / $347ScoredSep 25, 202624Qwen3.8 MaxQwenScore1472Score / $297ScoredSep 25, 202625GLM 5.2Z.ai753BScore1470Score / $1615Score / GB VRAM~3.1ScoredSep 25, 202626MiMo-V2.5-ProXiaomi1TScore1469Score / $2412Score / GB VRAM~2.3ScoredSep 25, 202627Qwen3.7 MaxQwenScore1466Score / $391ScoredSep 25, 202628Gemini 3.5 FlashGoogleScore1465Score / $326ScoredSep 25, 202629Grok 4.5xAIScore1462Score / $244ScoredSep 25, 202630Claude Sonnet 4.5AnthropicScore1462Score / $97ScoredSep 25, 202631GPT-5.6 TerraOpenAIScore1462Score / $244ScoredSep 25, 202632DeepSeek V4 ProDeepSeek1.6TScore1461Score / $933Score / GB VRAM~1.4ScoredSep 25, 202633GLM 5.1Z.ai754BScore1460Score / $482Score / GB VRAM~3.1ScoredSep 25, 202634GPT-5.4OpenAIScore1460Score / $195ScoredSep 25, 202635GPT-5.2 ChatOpenAIScore1457Score / $208ScoredSep 25, 202636Kimi K2.6Moonshot AI1.1TScore1456Score / $796Score / GB VRAM~2.2ScoredSep 25, 202637Claude Opus 4.1AnthropicScore1454Score / $19ScoredSep 25, 202638Gemma 4 31BGoogle31.3BScore1453Score / $4403Score / GB VRAM~74ScoredSep 25, 202639Grok 4.6xAIScore1451Score / $242ScoredSep 25, 202640Grok 4.20xAIScore1450Score / $580ScoredSep 25, 202641MiMo-V2.6-FlashXiaomiScore1449Score / $5449ScoredSep 25, 202642Qwen3.6 Max PreviewQwenScore1449Score / $235ScoredSep 25, 202643Qwen3.7 PlusQwenScore1447Score / $1130ScoredSep 25, 202644GPT-6 SolOpenAIScore1447Score / $289ScoredSep 25, 202645GLM 5Z.ai754BScore1446Score / $753Score / GB VRAM~3ScoredSep 25, 202646GPT-5.6 LunaOpenAIScore1446Score / $2410ScoredSep 25, 202647Hy3Tencent299BScore1446Score / $2738Score / GB VRAM~7.7ScoredSep 25, 202648Gemini 3.5 Flash LiteGoogleScore1444Score / $1155ScoredSep 25, 202649Grok 4.20 Multi-AgentxAIScore1444Score / $577ScoredSep 25, 202650GPT-6 LunaOpenAIScore1443Score / $5773ScoredSep 25, 202651Kimi K2 ThinkingMoonshot AI1.1TScore1439Score / $576Score / GB VRAM~2.2ScoredSep 13, 202652Gemini 2.5 ProGoogleScore1438Score / $288ScoredSep 25, 202653Gemma 4 26B A4BGoogle25.8BScore1437Score / $6533Score / GB VRAM~88ScoredSep 25, 202654Grok 4.7xAIScore1437Score / $239ScoredSep 25, 202655GPT-5.3 ChatOpenAIScore1436Score / $103ScoredSep 25, 202656Qwen3.6 PlusQwenScore1436Score / $736ScoredSep 25, 202657MiniMax M3MiniMax427BScore1435Score / $1495Score / GB VRAM~5.3ScoredSep 25, 202658Qwen3.5 397B A17BQwen403BScore1434Score / $613Score / GB VRAM~5.6ScoredSep 25, 202659GPT-5.4 MiniOpenAIScore1434Score / $637ScoredSep 25, 202660Kimi K2.5Moonshot AI1.1TScore1432Score / $636Score / GB VRAM~2.1ScoredSep 25, 202661MiMo-V2.5Xiaomi311BScore1430Score / $6010Score / GB VRAM~7.3ScoredSep 25, 202662DeepSeek V4 FlashDeepSeek291BScore1428Score / $17059Score / GB VRAM~7.8ScoredSep 25, 202663InklingThinking Machines952BScore1427Score / $352Score / GB VRAM~2.4ScoredSep 25, 202664GPT-5.1OpenAIScore1426Score / $285ScoredSep 25, 202665GLM 4.7Z.ai358BScore1425Score / $815Score / GB VRAM~6.3ScoredSep 25, 202666Qwen3 MaxQwenScore1424Score / $365ScoredSep 25, 202667Qwen3.8 27BQwen27.8BScore1423Score / $955Score / GB VRAM~81ScoredSep 25, 202668GLM 5V TurboZ.aiScore1423Score / $356ScoredSep 25, 202669Mistral Medium 3.5Mistral AI128BScore1421Score / $189Score / GB VRAM~18ScoredSep 25, 202670DeepSeek V3.2DeepSeek685BScore1420Score / $4587Score / GB VRAM~3.3ScoredSep 25, 202671Claude Haiku 4.5AnthropicScore1415Score / $283ScoredSep 25, 202672Claude Opus 4AnthropicScore1415Score / $19ScoredSep 25, 202673Grok 4.3xAIScore1415Score / $566ScoredSep 25, 202674GPT-5OpenAIScore1415Score / $141ScoredSep 25, 202675Qwen3 235B A22B Instruct 2507Qwen235BScore1414Score / $4041Score / GB VRAM~9.5ScoredSep 25, 202676DeepSeek V3.2 ExpDeepSeek685BScore1414Score / $3449Score / GB VRAM~3.3ScoredSep 25, 202677GLM 4.6Z.ai357BScore1412Score / $807Score / GB VRAM~6.3ScoredSep 25, 202678Muse Glimmer 30BMeta29.8BScore1412Score / $1284Score / GB VRAM~75ScoredSep 25, 202679Qwen3 VL 235B A22B InstructQwen236BScore1409Score / $1601Score / GB VRAM~9.5ScoredSep 25, 202680Gemini 3.1 Flash Lite PreviewGoogleScore1407Score / $1877ScoredSep 25, 202681o1OpenAIScore1407Score / $23ScoredSep 25, 202682MiniMax M2.7MiniMax229BScore1407Score / $1675Score / GB VRAM~9.8ScoredSep 25, 202683Qwen3.5-122B-A10BQwen125BScore1406Score / $676Score / GB VRAM~18ScoredSep 25, 202684GLM 4.5Z.ai358BScore1405Score / $638Score / GB VRAM~6.2ScoredSep 25, 202685GPT-4.1OpenAIScore1403Score / $175ScoredSep 25, 202686o3OpenAIScore1403Score / $175ScoredSep 25, 202687DeepSeek V3.1DeepSeek685BScore1402Score / $1476Score / GB VRAM~3.2ScoredSep 25, 202688Qwen3.5-27BQwen27.8BScore1400Score / $898Score / GB VRAM~80ScoredSep 25, 202689Mistral Large 3 2512Mistral AIScore1400Score / $933ScoredSep 25, 202690Gemini 2.5 FlashGoogleScore1400Score / $1120ScoredSep 25, 202691R1 0528DeepSeek685BScore1397Score / $650Score / GB VRAM~3.2ScoredSep 25, 202692Hy3 previewTencent299BScore1397Score / $2329Score / GB VRAM~7.4ScoredSep 25, 202693Claude Sonnet 4AnthropicScore1396Score / $93ScoredSep 25, 202694DeepSeek V3.1 TerminusDeepSeek685BScore1394Score / $1394Score / GB VRAM~3.2ScoredSep 25, 202695Inkling SmallThinking Machines266BScore1393Score / $1161Score / GB VRAM~8.3ScoredSep 25, 202696GPT-5.4 NanoOpenAIScore1387Score / $2218ScoredSep 25, 202697Qwen3.5-35B-A3BQwen36BScore1386Score / $1848Score / GB VRAM~61ScoredSep 25, 202698Qwen3 235B A22B Thinking 2507Qwen235BScore1385Score / $602Score / GB VRAM~9.3ScoredSep 25, 202699Step 3.5 FlashStepFun199BScore1384Score / $4613Score / GB VRAM~11ScoredSep 25, 2026100Qwen3 Coder 480B A35BQwen480BScore1384Score / $1384Score / GB VRAM~4.6ScoredSep 25, 2026
Showing 1–100 of 168 models on this board.