• CodingCoding scoreClaude Opus 5.5Anthropic72.8
  1. 2Claude Sonnet 5.570.1
  2. 3Claude Fable 5.169.4
• Agents & tool useAgents scoreClaude Sonnet 5.5Anthropic78.1
  1. 2GPT-6 Astra77.0
  2. 3Claude Opus 5.576.9
• ReasoningReasoning scoreGPT-6.1 SolOpenAI78.6
  1. 2Gemini 4 Argon78.0
  2. 3
• Long contextContext windowGrok 4.1 Fast ReasoningxAI2M tokens
  1. 2Grok 4 Fast Reasoning2M tokens
  2. 3GPT-6 Astra1.1M tokens
• VisionRespan IndexClaude Opus 5.5Anthropic72.4
  1. 2GPT-6 Astra70.5
  2. 3Gemini 4 Argon70.3
• Lowest costBlended priceGPT-5 nanoOpenAI$0.138/M
  1. 2Qwen3.5-Flash$0.175/M
  2. 3