Leaderboards
Benchmarks
Every score carries its origin.
Arena Maths
Arena Score on maths prompts — not a percentageProvenance100 of 163 modelsscores Aug 10, 2026 – Sep 25, 2026
1Claude Fable 5AnthropicScore1526Score / $31ScoredSep 13, 20262Claude Opus 5AnthropicScore1523Score / $61ScoredSep 25, 20263Gemini 3.8 FlashGoogleScore1522Score / $812ScoredSep 25, 20264Gemini 3.6 FlashGoogleScore1518Score / $810ScoredAug 10, 20265Claude Fable 5.1AnthropicScore1517Score / $30ScoredSep 25, 20266DeepSeek V4.1 FlashDeepSeek763BScore1509Score / $3773Score / GB VRAM~3.1ScoredSep 25, 20267Gemini 3.7 FlashGoogleScore1507Score / $804ScoredSep 25, 20268Claude Opus 4.6AnthropicScore1507Score / $60ScoredSep 25, 20269Muse Spark 1.3MetaScore1505Score / $354ScoredSep 25, 202610GLM 5.3 FlashZ.ai321BScore1501Score / $6006Score / GB VRAM~7.4ScoredSep 25, 202611GPT-5.5OpenAIScore1500Score / $100ScoredSep 25, 202612Kimi K3Moonshot AI2.8TScore1499Score / $153Score / GB VRAM~0.86ScoredSep 25, 202613GLM 5.3Z.aiScore1497Score / $1313ScoredSep 25, 202614Gemini 3.5 FlashGoogleScore1496Score / $332ScoredSep 25, 202615Qwen3.8 MaxQwenScore1495Score / $302ScoredSep 25, 202616GPT-5.6 SolOpenAIScore1494Score / $299ScoredSep 25, 202617Claude Opus 4.7AnthropicScore1492Score / $60ScoredSep 25, 202618Muse Spark 1.1MetaScore1489Score / $350ScoredSep 25, 202619Gemini 3.1 Pro Preview Custom ToolsGoogleScore1488Score / $248ScoredSep 25, 202620GLM 5.2Z.ai753BScore1487Score / $1634Score / GB VRAM~3.1ScoredSep 25, 202621GPT-6 AstraOpenAIScore1486Score / $59ScoredSep 25, 202622Qwen3.7 MaxQwenScore1485Score / $396ScoredSep 25, 202623InklingThinking Machines952BScore1482Score / $366Score / GB VRAM~2.5ScoredSep 25, 202624Hy3Tencent299BScore1482Score / $2807Score / GB VRAM~7.9ScoredSep 25, 202625Kimi K2.6Moonshot AI1.1TScore1480Score / $810Score / GB VRAM~2.2ScoredSep 25, 202626MiMo-V2.5-ProXiaomi1TScore1479Score / $2429Score / GB VRAM~2.3ScoredSep 25, 202627GPT-5.6 TerraOpenAIScore1478Score / $246ScoredSep 25, 202628Claude Opus 4.8AnthropicScore1477Score / $59ScoredSep 25, 202629Qwen3.6 Max PreviewQwenScore1476Score / $239ScoredSep 25, 202630GLM 5.1Z.ai754BScore1475Score / $487Score / GB VRAM~3.1ScoredSep 25, 202631GPT-5.6 LunaOpenAIScore1474Score / $2457ScoredSep 25, 202632Grok 4.5xAIScore1471Score / $245ScoredSep 25, 202633Kimi K2 ThinkingMoonshot AI1.1TScore1470Score / $588Score / GB VRAM~2.2ScoredSep 13, 202634Gemma 4 31BGoogle31.3BScore1469Score / $4452Score / GB VRAM~74ScoredSep 25, 202635MiMo-V2.6-FlashXiaomiScore1469Score / $5523ScoredSep 25, 202636Gemma 4 26B A4BGoogle25.8BScore1468Score / $6673Score / GB VRAM~90ScoredSep 25, 202637Qwen3.7 PlusQwenScore1466Score / $1146ScoredSep 25, 202638Claude Sonnet 4.6AnthropicScore1463Score / $98ScoredSep 25, 202639Claude Opus 4.5AnthropicScore1463Score / $59ScoredSep 25, 202640GPT-5.4OpenAIScore1460Score / $195ScoredSep 25, 202641DeepSeek V4 ProDeepSeek1.6TScore1457Score / $931Score / GB VRAM~1.4ScoredSep 25, 202642Grok 4.20 Multi-AgentxAIScore1455Score / $582ScoredSep 25, 202643GPT-5.2 ChatOpenAIScore1455Score / $208ScoredSep 25, 202644Qwen3.6 PlusQwenScore1455Score / $746ScoredSep 25, 202645Qwen3.5 397B A17BQwen403BScore1452Score / $621Score / GB VRAM~5.7ScoredSep 25, 202646GPT-6 SolOpenAIScore1450Score / $290ScoredSep 25, 202647Qwen3.8 27BQwen27.8BScore1450Score / $973Score / GB VRAM~83ScoredSep 25, 202648Grok 4.20xAIScore1449Score / $580ScoredSep 25, 202649Muse Glimmer 30BMeta29.8BScore1448Score / $1316Score / GB VRAM~77ScoredSep 25, 202650o3OpenAIScore1448Score / $181ScoredSep 25, 202651Inkling SmallThinking Machines266BScore1447Score / $1206Score / GB VRAM~8.6ScoredSep 25, 202652GLM 5Z.ai754BScore1443Score / $752Score / GB VRAM~3ScoredSep 25, 202653Kimi K2.5Moonshot AI1.1TScore1443Score / $641Score / GB VRAM~2.2ScoredSep 25, 202654Grok 4.6xAIScore1443Score / $240ScoredSep 25, 202655Gemini 3.5 Flash LiteGoogleScore1440Score / $1152ScoredSep 25, 202656Gemini 2.5 ProGoogleScore1439Score / $288ScoredSep 25, 202657GPT-5.4 MiniOpenAIScore1439Score / $639ScoredSep 25, 202658MiMo-V2.5Xiaomi311BScore1438Score / $6042Score / GB VRAM~7.3ScoredSep 25, 202659GLM 5V TurboZ.aiScore1437Score / $359ScoredSep 25, 202660Gemini 3.1 Flash Lite PreviewGoogleScore1435Score / $1914ScoredSep 25, 202661Qwen3 MaxQwenScore1434Score / $368ScoredSep 25, 202662MiniMax M3MiniMax427BScore1433Score / $1493Score / GB VRAM~5.3ScoredSep 25, 202663Claude Opus 4.1AnthropicScore1433Score / $19ScoredSep 25, 202664Hy3 previewTencent299BScore1431Score / $2385Score / GB VRAM~7.6ScoredSep 25, 202665GPT-5.3 ChatOpenAIScore1431Score / $102ScoredSep 25, 202666GPT-6 LunaOpenAIScore1430Score / $5721ScoredSep 25, 202667Mistral Medium 3.5Mistral AI128BScore1429Score / $190Score / GB VRAM~18ScoredSep 25, 202668GLM 4.7Z.ai358BScore1428Score / $816Score / GB VRAM~6.3ScoredSep 25, 202669Claude Sonnet 4.5AnthropicScore1427Score / $95ScoredSep 25, 202670DeepSeek V3.2DeepSeek685BScore1427Score / $4609Score / GB VRAM~3.3ScoredSep 25, 202671Qwen3.5-122B-A10BQwen125BScore1426Score / $686Score / GB VRAM~18ScoredSep 25, 202672DeepSeek V4 FlashDeepSeek291BScore1425Score / $17027Score / GB VRAM~7.8ScoredSep 25, 202673GPT-5.4 NanoOpenAIScore1425Score / $2279ScoredSep 25, 202674GPT-5.1OpenAIScore1423Score / $285ScoredSep 25, 202675Grok 4.7xAIScore1423Score / $237ScoredSep 25, 202676Qwen3.5-27BQwen27.8BScore1423Score / $912Score / GB VRAM~81ScoredSep 25, 202677MiniMax M2.7MiniMax229BScore1422Score / $1692Score / GB VRAM~9.9ScoredSep 25, 202678Grok 4.3xAIScore1419Score / $568ScoredSep 25, 202679o4 MiniOpenAIScore1418Score / $322ScoredSep 25, 202680GLM 4.6Z.ai357BScore1418Score / $810Score / GB VRAM~6.3ScoredSep 25, 202681Qwen3 235B A22B Instruct 2507Qwen235BScore1417Score / $4049Score / GB VRAM~9.6ScoredSep 25, 202682Qwen3 Next 80B A3B InstructQwen81.3BScore1414Score / $1813Score / GB VRAM~28ScoredSep 25, 202683DeepSeek V3.1DeepSeek685BScore1413Score / $1488Score / GB VRAM~3.3ScoredSep 25, 202684DeepSeek V3.2 ExpDeepSeek685BScore1413Score / $3446Score / GB VRAM~3.3ScoredSep 25, 202685R1 0528DeepSeek685BScore1412Score / $657Score / GB VRAM~3.3ScoredSep 25, 202686GLM 4.5Z.ai358BScore1412Score / $642Score / GB VRAM~6.2ScoredSep 25, 202687Qwen3 VL 235B A22B InstructQwen236BScore1409Score / $1602Score / GB VRAM~9.5ScoredSep 25, 202688GPT-5OpenAIScore1409Score / $141ScoredSep 25, 202689o1OpenAIScore1409Score / $23ScoredSep 25, 202690Step 3.5 FlashStepFun199BScore1408Score / $4692Score / GB VRAM~11ScoredSep 25, 202691GPT-5 MiniOpenAIScore1405Score / $1405ScoredSep 25, 202692o3 Mini HighOpenAIScore1405Score / $319ScoredSep 25, 202693Gemini 2.5 FlashGoogleScore1405Score / $1124ScoredSep 25, 202694Claude Opus 4AnthropicScore1404Score / $19ScoredSep 25, 202695Mistral Large 3 2512Mistral AIScore1401Score / $934ScoredSep 25, 202696Qwen3.5-FlashQwenScore1401Score / $5390ScoredSep 25, 202697Claude Haiku 4.5AnthropicScore1400Score / $280ScoredSep 25, 202698Qwen3 VL 235B A22B ThinkingQwen236BScore1399Score / $354Score / GB VRAM~9.4ScoredSep 25, 202699Qwen3 32BQwen32.8BScore1399Score / $4995Score / GB VRAM~68ScoredSep 25, 2026100Qwen3.5-35B-A3BQwen36BScore1398Score / $1864Score / GB VRAM~62ScoredSep 25, 2026
Showing 1–100 of 163 models on this board.