Leaderboards

Benchmarks

Every score carries its origin.

Reasoning374Coding337Math221Agentic375Chat636Speech778
Arena Code (WebDev)95Arena Coding168LiveBench Coding58LiveCodeBench16

LiveBench Coding

% average over LiveBench code generation and completion tasks
Provenance58 modelsscores Jun 25, 2026
1Claude Sonnet 5.5AnthropicScore91.37Score / $9ScoredJun 25, 20262Claude Opus 5.5AnthropicScore89.25Score / $4ScoredJun 25, 20263Claude Fable 5.1AnthropicScore86.38Score / $2ScoredJun 25, 20264Claude Fable 5AnthropicScore85.99Score / $2ScoredJun 25, 20265GPT-5.6 SolOpenAIScore83.94Score / $17ScoredJun 25, 20266GPT-5.2-CodexOpenAIScore83.62Score / $6ScoredJun 25, 20267GPT-5.6 LunaOpenAIScore82.92Score / $138ScoredJun 25, 20268Union AlphastealthScore82.15ScoredJun 25, 20269GPT-5.5OpenAIScore82.15Score / $5ScoredJun 25, 202610Claude Opus 4.7AnthropicScore82.09Score / $3ScoredJun 25, 202611Claude Opus 4.8AnthropicScore81.83Score / $3ScoredJun 25, 202612GPT-6 SolOpenAIScore81.77Score / $16ScoredJun 25, 202613Claude Opus 5AnthropicScore81.45Score / $3ScoredJun 25, 202614Kimi K3Moonshot AI2.8TScore81.45Score / $8Score / GB VRAM~0.046ScoredJun 25, 202615Muse Spark 1.3MetaScore81.06Score / $19ScoredJun 25, 202616Claude Sonnet 5AnthropicScore80.68Score / $8ScoredJun 25, 202617GPT-6 AstraOpenAIScore80.36Score / $3ScoredJun 25, 202618DeepSeek V4.1 FlashDeepSeek763BScore80.04Score / $200Score / GB VRAM~0.17ScoredJun 25, 202619GLM 5.2Z.ai753BScore79.65Score / $88Score / GB VRAM~0.17ScoredJun 25, 202620Claude Opus 4.5AnthropicScore79.65Score / $3ScoredJun 25, 202621Claude Sonnet 4.6AnthropicScore79.27Score / $5ScoredJun 25, 202622GLM 5.3 FlashZ.ai321BScore78.95Score / $316Score / GB VRAM~0.39ScoredJun 25, 202623GLM 5.3Z.aiScore78.95Score / $69ScoredJun 25, 202624GPT-6 LunaOpenAIScore78.95Score / $316ScoredJun 25, 202625Gemini 3.7 FlashGoogleScore78.89Score / $42ScoredJun 25, 202626Kimi K2.6Moonshot AI1.1TScore78.57Score / $43Score / GB VRAM~0.12ScoredJun 25, 202627GPT-5.6 TerraOpenAIScore78.25Score / $13ScoredJun 25, 202628Qwen3.6 PlusQwenScore78.18Score / $40ScoredJun 25, 202629Claude Opus 4.6AnthropicScore78.18Score / $3ScoredJun 25, 202630Gemini 3.5 FlashGoogleScore78.18Score / $17ScoredJun 25, 202631Gemini 3.6 FlashGoogleScore77.86Score / $42ScoredJun 25, 202632Muse Spark 1.2MetaScore77.54Score / $18ScoredJun 25, 202633GPT-5.4OpenAIScore77.54Score / $10ScoredJun 25, 202634Muse Spark 1.1MetaScore77.16Score / $18ScoredJun 25, 202635Grok 4.7xAIScore77.16Score / $13ScoredJun 25, 202636Grok 4.6xAIScore76.78Score / $13ScoredJun 25, 202637Gemini 3.5 Flash LiteGoogleScore76.07Score / $61ScoredJun 25, 202638GPT-5.2 ChatOpenAIScore76.07Score / $11ScoredJun 25, 202639Ox AlphastealthScore75.75ScoredJun 25, 202640Qwen3.8 27BQwen27.8BScore75.69Score / $51Score / GB VRAM~4.3ScoredJun 25, 202641Qwen3.7 MaxQwenScore74.22Score / $20ScoredJun 25, 202642Kimi K2.7 CodeMoonshot AI1.1TScore73.96Score / $25Score / GB VRAM~0.11ScoredJun 25, 202643Qwen3 MaxQwenScore72.87Score / $19ScoredJun 25, 202644Qwen3.8 MaxQwenScore72.87Score / $15ScoredJun 25, 202645Qwen3.8 FlashQwenScore72.55Score / $190ScoredJun 25, 202646Gemini 3.8 FlashGoogleScore72.49Score / $39ScoredJun 25, 202647Qwen3.6 27BQwen27.8BScore71.79Score / $36Score / GB VRAM~4.1ScoredJun 25, 202648GPT-5.4 MiniOpenAIScore71.62Score / $32ScoredJun 25, 202649InklingThinking Machines952BScore71.02Score / $18Score / GB VRAM~0.12ScoredJun 25, 202650GPT-5.4 NanoOpenAIScore70.84Score / $113ScoredJun 25, 202651Nemotron 3 UltraNVIDIA561BScore70.7Score / $32Score / GB VRAM~0.2ScoredJun 25, 202652DeepSeek V4 ProDeepSeek1.6TScore69.99Score / $45Score / GB VRAM~0.069ScoredJun 25, 202653Grok 4.3xAIScore69.93Score / $28ScoredJun 25, 202654DeepSeek V4 FlashDeepSeek291BScore69.23Score / $827Score / GB VRAM~0.38ScoredJun 25, 202655Grok 4.5xAIScore68.59Score / $11ScoredJun 25, 202656MiniMax M3MiniMax427BScore68.2Score / $71Score / GB VRAM~0.25ScoredJun 25, 202657DeepSeek V4 Flash Vision ExpDeepSeek305BScore68.2Score / $105Score / GB VRAM~0.36ScoredJun 25, 202658Grok Build 0.1xAIScore65.39Score / $33ScoredJun 25, 2026