Leaderboards

Benchmarks

Every score carries its origin.

Reasoning374Coding337Math221Agentic375Chat636Speech778
Arena Agent55Arena Agent · Recovery55Arena Agent · Steerability55Arena Agent · Task outcome55Arena Agent · Tool use55LiveBench Agentic Coding58SWE-bench Verified42

SWE-bench Verified

% of real GitHub issues resolved end-to-end
Provenance42 modelsscores Apr 2, 2024 – Feb 19, 2026
1Claude Opus 4.5AnthropicScore79.2Score / $3ScoredDec 15, 20252MiniMax M2.5MiniMax229BScore75.8Score / $80Score / GB VRAM~0.53ScoredFeb 17, 20263Claude Opus 4.6AnthropicScore75.6Score / $3ScoredFeb 17, 20264Claude Sonnet 4.5AnthropicScore74.8Score / $5ScoredNov 3, 20255GPT-5.5OpenAIScore74.4Score / $5ScoredOct 15, 20256Claude Opus 4AnthropicScore73.2Score / $1ScoredMay 22, 20257GPT-5.2-CodexOpenAIScore72.8Score / $5ScoredFeb 19, 20268GLM 5Z.ai754BScore72.8Score / $38Score / GB VRAM~0.15ScoredFeb 17, 20269Kimi K2.5Moonshot AI1.1TScore70.8Score / $31Score / GB VRAM~0.11ScoredFeb 17, 202610DeepSeek V3.2DeepSeek685BScore70Score / $226Score / GB VRAM~0.16ScoredFeb 17, 202611Qwen3 Coder 480B A35BQwen480BScore69.6Score / $70Score / GB VRAM~0.23ScoredAug 5, 202512GPT-5.2 ChatOpenAIScore69Score / $10ScoredDec 11, 202513GLM 4.6Z.ai357BScore68.2Score / $39Score / GB VRAM~0.3ScoredSep 30, 202514Claude Haiku 4.5AnthropicScore66.6Score / $13ScoredFeb 17, 202615GPT-5.1-CodexOpenAIScore66Score / $7ScoredNov 24, 202516GPT-5.1OpenAIScore66Score / $13ScoredNov 20, 202517Kimi K2 0711Moonshot AI1TScore65.4Score / $28Score / GB VRAM~0.1ScoredJul 16, 202518GPT-5OpenAIScore65Score / $7ScoredAug 7, 202519Claude Sonnet 4AnthropicScore64.9Score / $4ScoredJul 26, 202520GLM 4.5Z.ai358BScore64.2Score / $29Score / GB VRAM~0.28ScoredJul 28, 202521Kimi K2 ThinkingMoonshot AI1.1TScore63.4Score / $25Score / GB VRAM~0.095ScoredDec 10, 202522MiniMax M2MiniMax229BScore61Score / $60Score / GB VRAM~0.42ScoredNov 24, 202523Qwen3 Coder 30B A3B InstructQwen30.5BScore60.4Score / $224Score / GB VRAM~3.1ScoredSep 1, 202524o3OpenAIScore58.4Score / $7ScoredJul 26, 202525GPT-5 MiniOpenAIScore56.2Score / $56ScoredFeb 17, 202626Devstral 2 2512Mistral AI125BScore53.8Score / $27Score / GB VRAM~0.68ScoredDec 9, 202527Gemini 2.5 ProGoogleScore53.6Score / $11ScoredJul 26, 202528o4 MiniOpenAIScore45Score / $10ScoredJul 26, 202529Nova Premier 1.0AmazonScore42.4Score / $3ScoredMay 27, 202530o3 MiniOpenAIScore42.4Score / $10ScoredFeb 14, 202531GPT-4.1OpenAIScore39.6Score / $5ScoredJul 26, 202532GPT-4o (2024-05-13)OpenAIScore38.8Score / $3ScoredOct 28, 202433GPT-5 NanoOpenAIScore34.8Score / $174ScoredAug 7, 202534GPT-4oOpenAIScore32.6Score / $3ScoredJun 12, 202435Gemini 2.5 FlashGoogleScore28.7Score / $23ScoredJul 26, 202536gpt-oss-120bOpenAI120BScore26Score / $153Score / GB VRAM~0.34ScoredAug 7, 202537GPT-4.1 MiniOpenAIScore23.9Score / $15ScoredJul 20, 202538GPT-4OpenAIScore22.4Score / $0ScoredApr 2, 202439GPT-4o (2024-11-20)OpenAIScore21.6Score / $2ScoredJul 20, 202540Llama 4 MaverickMeta402BScore21Score / $32Score / GB VRAM~0.083ScoredJul 20, 202541Llama 4 ScoutMeta109BScore9.1Score / $30Score / GB VRAM~0.13ScoredJul 20, 202542Qwen2.5 Coder 32B InstructQwen32.8BScore9Score / $9Score / GB VRAM~0.44ScoredAug 3, 2025