Model Markets
Get Free API KeyAPI

Rankings

Intelligence, cost, and their combined efficiency—using every eligible model.

Intelligence

22 eligible · v1.2 · Sep 3, 2026 · Higher is better

  1. 96.1Claude Fable 5
  2. 94.1Claude Opus 5
  3. 87.8Gemini 3.7 Flash
  4. 85.2GPT-5.6 Sol
  5. 84.2GPT-5.5
  6. 78.9Claude Opus 4.7
  7. 76.2Claude Opus 4.6
  8. 75.0GPT-5.6 Terra
  9. 72.8GPT-5.4
  10. 72.0Gemini 3.5 Flash
  11. 71.3Kimi K3
  12. 69.9Claude Opus 4.8
  13. 67.6Grok 4.6
  14. 66.5Gemini 3.6 Flash
  15. 60.5Grok 4.5
  16. 60.3Claude Sonnet 4.6
  17. 52.8GPT-5.6 Luna
  18. 49.3GPT-5.2
  19. 41.8DeepSeek V4 Flash
  20. 20.2GPT-5.4 Mini
  21. 17.8Gemini 3.5 Flash Lite
  22. 14.7GPT-5.4 Nano
Equal thirds: LiveBench overall, ARC-AGI-2 verified, and LMArena Text overall. All three are required.

Cost

36 eligible · output cost per LiveBench case · Lower is better

  1. $0.0020Grok Build 0.1
  2. $0.0060DeepSeek V4 Flash
  3. $0.0087GLM 5.3 Flash
  4. $0.017MiniMax M3
  5. $0.026GPT-5.6 Luna
  6. $0.028Grok 4.3
  7. $0.029Gemini 3.5 Flash Lite
  8. $0.035DeepSeek V4 Flash Vision Exp
  9. $0.054DeepSeek V4 Pro
  10. $0.055Gemini 3.6 Flash
  11. $0.056GLM 5.2
  12. $0.057Qwen3.7 Max
  13. $0.058GPT-5.4 Nano
  14. $0.063Grok 4.5
  15. $0.085Gemini 3.7 Flash
  16. $0.086Qwen3.8 27B
  17. $0.095Kimi K2.6
  18. $0.097Grok 4.6
  19. $0.107Qwen3.8 Max
  20. $0.117GPT-5.6 Sol
  21. $0.140Gemini 3.5 Flash
  22. $0.161Gemini 3.1 Pro
  23. $0.161GPT-5.2
  24. $0.194Kimi K3
  25. $0.198Claude Sonnet 4.6
  26. $0.214GPT-5.4 Mini
  27. $0.241Claude Opus 4.6
  28. $0.248GLM 5.3
  29. $0.266GPT-5.6 Terra
  30. $0.274GPT-5.4
  31. $0.286Claude Opus 4.7
  32. $0.341GPT-5.5
  33. $0.371Claude Opus 5
  34. $0.380Claude Opus 4.5
  35. $0.604Claude Opus 4.8
  36. $1.01Claude Fable 5
Published output tokens per LiveBench case multiplied by the lowest current sourced USD output-token rate.

Efficiency

22 eligible · v1.0 · Higher is better

  1. 70.9DeepSeek V4 Flash
  2. 68.1Gemini 3.7 Flash
  3. 63.6GPT-5.6 Sol
  4. 62.0GPT-5.6 Luna
  5. 61.6Gemini 3.6 Flash
  6. 57.3Grok 4.5
  7. 56.8Claude Opus 5
  8. 56.6Grok 4.6
  9. 55.3Gemini 3.5 Flash
  10. 52.7GPT-5.5
  11. 52.1Claude Opus 4.6
  12. 51.8Claude Opus 4.7
  13. 51.7Kimi K3
  14. 50.5GPT-5.6 Terra
  15. 49.1GPT-5.4
  16. 48.0Claude Fable 5
  17. 46.0Claude Sonnet 4.6
  18. 43.6Gemini 3.5 Flash Lite
  19. 42.6GPT-5.2
  20. 40.0Claude Opus 4.8
  21. 35.3GPT-5.4 Nano
  22. 25.3GPT-5.4 Mini
Equal halves: MM Relative Intelligence and logarithmically normalized affordability across the complete common cohort.

Every panel includes the complete eligible cohort in exact rank order. Scroll horizontally to compare all models; unavailable benchmark, token, or pricing inputs remain unranked rather than estimated.

Send Feedback