The best LLMs in the overall index

Overall index from published benchmarks. 100 is the average of a fixed reference group of models, higher is better.

  1. Claude Fable 5.2 Anthropic131.0
  2. Gemini 4 Carbon Google DeepMind127.0
  3. Claude Opus 5.5 Anthropic125.7
  4. Claude Fable 5.1 Anthropic124.1
  5. Claude Sonnet 5.5 Anthropic121.8
  6. Gemini 4 Argon Google DeepMind121.4
  7. GPT-6 Astra OpenAI120.2
  8. Muse Wonder Full Meta118.1
  9. GPT-6.1 Sol OpenAI118.0
  10. Claude Opus 5 Anthropic117.4
  11. Muse Wonder Meta113.4
  12. Claude Fable 5 Anthropic112.8
  13. Kimi K3.1 Pro Moonshot AI112.6
  14. Muse Ember 1.0 Meta110.4
  15. GPT-6 Sol OpenAI109.1
  16. GPT-5.6 Sol OpenAI109.1
  17. Gemini 3.7 Flash Google107.9
  18. Muse Spark 1.3 Meta107.3
  19. Gemini 3.8 Flash Google106.8
  20. DeepSeek V4.1 Flash 1017 DeepSeek106.7