Leaderboards

Benchmarks

Every score carries its origin.

Reasoning374Coding337Math221Agentic375Chat636Speech778
Arena Hard Prompts168GPQA Diamond16LiveBench58LiveBench Data Analysis58LiveBench Reasoning58MMLU-Pro16

LiveBench

Average % across monthly-refreshed reasoning/coding/math tasks
Provenance58 modelsscores Jun 25, 2026
1Claude Fable 5.1AnthropicScore83.41Score / $2ScoredJun 25, 20262Claude Opus 5.5AnthropicScore83.22Score / $4ScoredJun 25, 20263Claude Fable 5AnthropicScore82.97Score / $2ScoredJun 25, 20264GPT-6 AstraOpenAIScore82.16Score / $3ScoredJun 25, 20265Muse Spark 1.3MetaScore81.59Score / $19ScoredJun 25, 20266DeepSeek V4.1 FlashDeepSeek763BScore81.11Score / $203Score / GB VRAM~0.17ScoredJun 25, 20267GPT-5.6 SolOpenAIScore81.05Score / $16ScoredJun 25, 20268GPT-5.5OpenAIScore80.19Score / $5ScoredJun 25, 20269Claude Opus 5AnthropicScore80.08Score / $3ScoredJun 25, 202610GPT-6 SolOpenAIScore79.25Score / $16ScoredJun 25, 202611Kimi K3Moonshot AI2.8TScore79.19Score / $8Score / GB VRAM~0.045ScoredJun 25, 202612Gemini 3.7 FlashGoogleScore78.83Score / $42ScoredJun 25, 202613Qwen3.8 MaxQwenScore78.46Score / $16ScoredJun 25, 202614Qwen3 MaxQwenScore78.46Score / $20ScoredJun 25, 202615Grok 4.6xAIScore78.04Score / $13ScoredJun 25, 202616GPT-5.4OpenAIScore77.97Score / $10ScoredJun 25, 202617Muse Spark 1.2MetaScore77.95Score / $18ScoredJun 25, 202618GPT-5.6 TerraOpenAIScore77.94Score / $13ScoredJun 25, 202619Claude Sonnet 5.5AnthropicScore77.75Score / $8ScoredJun 25, 202620Grok 4.7xAIScore77.4Score / $13ScoredJun 25, 202621DeepSeek V4 Flash Vision ExpDeepSeek305BScore76.76Score / $119Score / GB VRAM~0.4ScoredJun 25, 202622Claude Opus 4.7AnthropicScore76.53Score / $3ScoredJun 25, 202623Claude Opus 4.8AnthropicScore76.22Score / $3ScoredJun 25, 202624Qwen3.8 FlashQwenScore76.19Score / $199ScoredJun 25, 202625GLM 5.3Z.aiScore76.14Score / $67ScoredJun 25, 202626Union AlphastealthScore76.13ScoredJun 25, 202627Claude Sonnet 5AnthropicScore76.04Score / $8ScoredJun 25, 202628Gemini 3.8 FlashGoogleScore75.83Score / $40ScoredJun 25, 202629Grok 4.5xAIScore75.77Score / $13ScoredJun 25, 202630Muse Spark 1.1MetaScore75.3Score / $18ScoredJun 25, 202631Qwen3.8 27BQwen27.8BScore75.27Score / $51Score / GB VRAM~4.3ScoredJun 25, 202632Gemini 3.5 FlashGoogleScore74.64Score / $17ScoredJun 25, 202633GPT-5.2 ChatOpenAIScore74.63Score / $11ScoredJun 25, 202634Claude Opus 4.6AnthropicScore74.52Score / $3ScoredJun 25, 202635GPT-5.2-CodexOpenAIScore73.98Score / $5ScoredJun 25, 202636Gemini 3.6 FlashGoogleScore73.59Score / $39ScoredJun 25, 202637GPT-5.6 LunaOpenAIScore73.56Score / $123ScoredJun 25, 202638GLM 5.2Z.ai753BScore73.16Score / $80Score / GB VRAM~0.15ScoredJun 25, 202639Qwen3.7 MaxQwenScore73.14Score / $20ScoredJun 25, 202640Claude Sonnet 4.6AnthropicScore72.99Score / $5ScoredJun 25, 202641Claude Opus 4.5AnthropicScore72.58Score / $3ScoredJun 25, 202642GPT-6 LunaOpenAIScore72.03Score / $288ScoredJun 25, 202643InklingThinking Machines952BScore71.92Score / $18Score / GB VRAM~0.12ScoredJun 25, 202644GLM 5.3 FlashZ.ai321BScore71.59Score / $286Score / GB VRAM~0.35ScoredJun 25, 202645DeepSeek V4 ProDeepSeek1.6TScore71.57Score / $46Score / GB VRAM~0.071ScoredJun 25, 202646Kimi K2.6Moonshot AI1.1TScore70.54Score / $39Score / GB VRAM~0.11ScoredJun 25, 202647GPT-5.4 NanoOpenAIScore69.58Score / $111ScoredJun 25, 202648Ox AlphastealthScore69.24ScoredJun 25, 202649Qwen3.6 PlusQwenScore68.91Score / $35ScoredJun 25, 202650Kimi K2.7 CodeMoonshot AI1.1TScore68.41Score / $23Score / GB VRAM~0.1ScoredJun 25, 202651Grok Build 0.1xAIScore67.78Score / $34ScoredJun 25, 202652Nemotron 3 UltraNVIDIA561BScore67.36Score / $31Score / GB VRAM~0.19ScoredJun 25, 202653MiniMax M3MiniMax427BScore67.26Score / $70Score / GB VRAM~0.25ScoredJun 25, 202654GPT-5.4 MiniOpenAIScore66.37Score / $29ScoredJun 25, 202655DeepSeek V4 FlashDeepSeek291BScore65.48Score / $782Score / GB VRAM~0.36ScoredJun 25, 202656Qwen3.6 27BQwen27.8BScore64.03Score / $32Score / GB VRAM~3.7ScoredJun 25, 202657Gemini 3.5 Flash LiteGoogleScore63.94Score / $51ScoredJun 25, 202658Grok 4.3xAIScore62.25Score / $25ScoredJun 25, 2026