Skip to content

Models

114 models that appear on two or more of the boards we track. Each page lists every result as published. No overall score.

Claude Code AgentAnthropicCode1 result Claude Fable 5AnthropicCodeAgenticWebsitesDesignAndroid9 results Claude Fable 5.1AnthropicCodeAgenticWebsitesDesignAndroid11 results Claude Haiku 4.5AnthropicCode1 result Claude Haiku 5.5AnthropicWebsitesDesign3 results Claude Opus 4.5AnthropicCodeWebsites2 results Claude Opus 4.6AnthropicCodeWebsitesDesign4 results Claude Opus 4.6 ThinkingAnthropicDesign1 result Claude Opus 4.7AnthropicWebsitesAndroid2 results Claude Opus 4.8AnthropicCodeAgenticWebsitesAndroid6 results Claude Opus 5AnthropicCodeAgenticWebsitesDesignAndroid14 results Claude Opus 5.5AnthropicAgenticWebsitesDesignAndroid5 results Claude Sonnet 4.5AnthropicCode1 result Claude Sonnet 4.6AnthropicWebsitesDesignAndroid3 results Claude Sonnet 5AnthropicCodeAgenticWebsitesDesignAndroid8 results Claude Sonnet 5.5AnthropicAgenticWebsitesDesign5 results DeepSeek V3.2DeepSeekCode2 results DeepSeek V3.2 ExpDeepSeekCode1 result DeepSeek V3.2 ThinkingDeepSeekWebsites1 result DeepSeek V4 FlashDeepSeekWebsitesDesignAndroid3 results DeepSeek V4 Flash 0731DeepSeekCodeDesign3 results DeepSeek V4 ProDeepSeekCodeWebsitesDesignAndroid4 results DeepSeek V4 Pro 0813DeepSeekCodeWebsites4 results DeepSeek V4.1 FlashDeepSeekWebsitesDesign3 results DeepSeek V4.1 Flash 1017DeepSeekWebsitesDesign3 results Devstral 2MistralCodeWebsites3 results Devstral Small 2MistralCode1 result Devstral Small 2507MistralDesign1 result Gemini 2.5 Pro Preview 05 06GoogleCode1 result Gemini 2.5 Pro Preview 06 05GoogleCode2 results Gemini 3 FlashGoogleCode1 result Gemini 3 Flash PreviewGoogleWebsitesAndroid2 results Gemini 3 ProGoogleCode1 result Gemini 3.1 ProGoogleCode1 result Gemini 3.1 Pro PreviewGoogleWebsitesAndroid2 results Gemini 3.5 FlashGoogleWebsitesAndroid2 results Gemini 3.5 Flash LiteGoogleAndroid1 result Gemini 3.6 FlashGoogleWebsitesDesignAndroid3 results Gemini 3.7 FlashGoogleCodeAgenticWebsitesDesignAndroid7 results Gemini 3.8 FlashGoogleCodeAgenticWebsitesDesignAndroid8 results Gemini 4 ArgonGoogleWebsitesDesign3 results Gemma 3 27BGoogleCode1 result Gemma 4 26B A4BGoogleWebsites1 result Gemma 4 31BGoogleWebsitesAndroid2 results GLM-4.6Z.aiCode1 result GLM-5Z.aiCodeWebsites3 results GLM-5.1Z.aiWebsitesAndroid3 results GLM-5.2Z.aiCodeWebsitesDesignAndroid6 results GLM-5.3Z.aiCodeAgenticWebsites5 results GLM-5.3 FlashZ.aiCodeAgenticWebsites5 results GLM-5v TurboZ.aiWebsites1 result GPT Oss 120BOpenAICodeDesign4 results GPT Oss 20BOpenAICode1 result GPT-5OpenAICode3 results GPT-5 MiniOpenAICode1 result GPT-5.2OpenAICode1 result GPT-5.2 CodexOpenAICode1 result GPT-5.4OpenAICodeAndroid2 results GPT-5.5OpenAIWebsitesAndroid2 results GPT-5.6 LunaOpenAICodeAgenticAndroid4 results GPT-5.6 SolOpenAICodeAgenticWebsitesDesignAndroid10 results GPT-5.6 TerraOpenAICodeAgenticWebsitesAndroid4 results GPT-6 AstraOpenAICodeAgenticWebsitesDesignAndroid11 results GPT-6 LunaOpenAIAgenticWebsitesDesignAndroid6 results GPT-6 SolOpenAIAgenticWebsitesDesignAndroid6 results GPT-6.1 SolOpenAIAgenticWebsitesDesign4 results Grok 4SpaceXAICodeDesign2 results Grok 4.20 0309SpaceXAIWebsites1 result Grok 4.3SpaceXAICodeWebsitesDesign6 results Grok 4.5SpaceXAICodeAgenticWebsites6 results Grok 4.6SpaceXAICodeAgenticWebsites5 results Grok 4.7SpaceXAIAgenticWebsitesDesign4 results Grok Build 0.1SpaceXAIWebsites1 result Grok Code Fast 1SpaceXAIWebsitesDesign2 results Kimi K2 InstructMoonshotCode1 result Kimi K2.5MoonshotCodeWebsites2 results Kimi K2.6MoonshotWebsitesDesignAndroid4 results Kimi K2.7 CodeMoonshotWebsitesAndroid3 results Kimi K3MoonshotCodeWebsitesDesignAndroid8 results Kimi K3.1 ProMoonshotWebsites1 result Llama 3.1 Nemotron Ultra 253BVendor not listedDesign1 result Llama 3.3 Instruct TurboVendor not listedCode1 result Llama 4 MaverickVendor not listedCodeDesign2 results Llama 4 ScoutVendor not listedCodeDesign2 results Llama3.1 405B InstructVendor not listedCode1 result Llama4 Maverick 17B InstructVendor not listedCode1 result MiniMax 2.1MiniMaxCode1 result MiniMax M2.5MiniMaxCode1 result MiniMax M2.7MiniMaxWebsitesAndroid2 results MiniMax M3MiniMaxCodeWebsitesDesignAndroid8 results MiniMax M3.1 FlashMiniMaxWebsites2 results Mistral Large 3MistralCodeWebsites3 results Mistral Large 4MistralWebsites2 results Mistral Medium 3.5MistralCodeWebsites3 results Muse GlimmerMetaWebsites1 result Muse SparkMetaCodeWebsites3 results Muse Spark 1.1MetaCodeWebsites4 results Muse Spark 1.2MetaCodeWebsitesDesign5 results Muse Spark 1.3MetaAgenticWebsitesDesign9 results Muse WonderMetaWebsitesDesign3 results Muse Wonder FullMetaWebsites2 results Qwen 3 235B A22BAlibabaCode1 result Qwen 3 Coder 480B A35BAlibabaCode1 result Qwen 3 Coder 480B A35B InstructAlibabaWebsites1 result Qwen 3.5 397B A17BAlibabaWebsites1 result Qwen 3.6 27BAlibabaCodeAndroid3 results Qwen 3.6 35B A3BAlibabaCodeAndroid2 results Qwen 3.7 MaxAlibabaWebsitesDesignAndroid3 results Qwen 3.7 PlusAlibabaAndroid1 result Qwen 3.8 27BAlibabaCodeWebsites4 results Qwen 3.8 Flash NextAlibabaWebsites1 result Qwen 3.8 MaxAlibabaCodeWebsitesAndroid7 results Qwen 3.8 Max 0902AlibabaAgenticWebsites3 results Qwen 3.8 Max PreviewAlibabaDesign1 result

Grouped by model version; effort settings like (max) or (xhigh) stay on each result as the board lists them.

Leaderboard historyTracking since Oct 16, 2026. Nothing is deleted.

Uncensored and low-refusal models12 models on the UGI benchmark, where to run them and the VRAM they need.