OpenAI AI models: benchmarks and ranking

OpenAI — provider rank: #2, overall score: 85.2, models: 54, top model: GPT-6 Astra.

More from OpenAI

  1. GPT-6 Astra — #2 · Score 92.2
  2. GPT-5.6 Sol — #7 · Score 88.0
  3. GPT-5.5 Pro — #8 · Score 86.9
  4. GPT-6 Sol — #9 · Score 86.5
  5. GPT-5.4 Pro — #10 · Score 85.2
  6. GPT-5.5 — #14 · Score 84.4
  7. GPT-5.4 — #19 · Score 82.1
  8. GPT-5.6 Terra — #21 · Score 82.1
  9. GPT-5.6 Luna — #36 · Score 73.8
  10. GPT-6 Luna — #39 · Score 73.5
  11. GPT-5.3 Codex — #41 · Score 70.0
  12. GPT-5.2.11 — #42 · Score 69.9
  13. GPT-5.2 Pro — #44 · Score 69.5
  14. GPT-5.2 — #49 · Score 65.2
  15. GPT-5.1 — #59 · Score 54.0
  16. GPT-5 Pro — #63 · Score 52.6
  17. GPT-5.4 nano — #67 · Score 47.0
  18. GPT-5.4 mini — #69 · Score 46.4
  19. GPT-5 — #70 · Score 46.2
  20. o3-pro — #75 · Score 43.8
  21. o3 — #76 · Score 43.4
  22. o4-mini — #88 · Score 38.7
  23. GPT-5 mini — #92 · Score 34.1
  24. GPT-5 nano — #99 · Score 30.4
  25. o1 — #104 · Score 27.8
  26. GPT-4.1 mini — #106 · Score 26.8
  27. GPT-4.5 — #108 · Score 25.4
  28. o3-mini — #109 · Score 25.3
  29. GPT-4.1 — #111 · Score 24.8
  30. GPT-4.1 nano — #112 · Score 24.3
RankProviderOverall scoreTop modelModels
186.7Claude Opus 5.527
285.2GPT-6 Astra54
377.7Muse Spark 1.323
474.4Gemini 3.8 Flash33
572.7Grok 4.713
670.1MiMo-V2.6-Pro4
769.8Kimi K35
868.1Qwen3.8 Max46
967.8DeepSeek V4.1 Flash22
1066.0GLM-5.310
1153.3—2
1242.0Inkling2
1340.3MiniMax-M35
1432.7—5
1529.6Nemotron 3 Ultra 550B A55B9
1617.6Mistral 3.526
1716.7Phi-46
1815.8—3
1913.3Mercury 2.51
2012.9Command R3

A provider’s result on each site is that of its best model there.