Ir para o conteúdo
AI.info

Model Tracker

SimpleQA Verified leaderboard

The models in the Model Tracker that Epoch AI has measured on SimpleQA Verified, ranked by the highest score of each: the name Epoch AI gives it, the setting it ran and the day.

Em inglês

The models in the Model Tracker that Epoch AI has measured on SimpleQA Verified, ranked by the highest score of each: the name Epoch AI gives it, the setting it ran and the day. Epoch AI, ‘Capabilities & benchmarking’, epoch.ai/benchmarks, CC BY 4.0: Epoch AI’s own runs and its Capabilities Index.

  1. 1. Claude Fable 5.2: 94.0% (± 0.8) — claude-fable-5-2_max, max, 6 November 2026
  2. 2. Gemini 4 Argon: 76.0% (± 1.4) — gemini-4-argon_high, high, 13 October 2026
  3. 3. GPT-6 Astra: 75.6% (± 1.4) — gpt-6-astra_max, max, 30 August 2026
  4. 4. GPT-6.1 Sol: 73.9% (± 1.4) — gpt-6.1-sol_max, max, 29 September 2026
  5. 5. Gemini 3.1 Pro: 73.5% (± 1.4) — gemini-3.1-pro-preview_high, high, 10 August 2026
  6. 6. Claude Opus 5.5: 72.2% (± 1.4) — claude-opus-5-5_max, max, 22 September 2026
  7. 7. Claude Fable 5.1: 70.8% (± 1.4) — claude-fable-5-1_max, max, 1 September 2026
  8. 8. Claude Fable 5: 70.7% (± 1.4) — claude-fable-5_xhigh, xhigh, 10 August 2026
  9. 9. Gemini 3.8 Flash: 69.7% (± 1.5) — gemini-3.8-flash_high, high, 2 September 2026
  10. 9. GPT-5.6 Sol: 69.7% (± 1.5) — gpt-5.6-sol_max, max, 10 August 2026
  11. 11. Gemini 3.7 Flash: 69.2% (± 1.5) — gemini-3.7-flash_high, high, 27 August 2026
  12. 12. Gemini 3 Flash: 66.8% (± 1.5) — gemini-3-flash-preview_high, high, 27 August 2026
  13. 13. Gemini 3.5 Flash: 66.2% (± 1.5) — gemini-3.5-flash_high, high, 27 August 2026
  14. 13. Gemini 3.6 Flash: 66.2% (± 1.5) — gemini-3.6-flash_high, high, 27 August 2026
  15. 15. GPT-5.5: 63.0% (± 1.5) — gpt-5.5_xhigh, xhigh, 27 August 2026
  16. 16. GPT-6 Sol: 60.7% (± 1.5) — gpt-6-sol_max, max, 22 September 2026
  17. 17. Claude Opus 5: 59.9% (± 1.6) — claude-opus-5_max, max, 10 August 2026
  18. 18. Grok 4.7: 56.0% (± 1.6) — grok-4.7_xhigh, xhigh, 22 September 2026
  19. 19. Claude Opus 4.8: 53.0% (± 1.6) — claude-opus-4-8_max, max, 27 August 2026
  20. 20. Claude Opus 4.7: 51.7% (± 1.6) — claude-opus-4-7_xhigh, xhigh, 27 August 2026
  21. 21. GPT-5: 50.1% (± 1.6) — gpt-5-2025-08-07_high, high, 27 August 2026
  22. 22. o3: 49.4% (± 1.6) — o3-2025-04-16_high, high, 27 August 2026
  23. 23. Grok 4.6: 49.3% (± 1.6) — grok-4.6_high, high, 27 August 2026
  24. 24. Grok 4.5: 48.3% (± 1.6) — grok-4.5_high, high, 27 August 2026
  25. 25. GPT-5.1: 48.0% (± 1.6) — gpt-5.1-2025-11-13_high, high, 27 August 2026
  26. 26. Claude Opus 4.6: 47.0% (± 1.6) — claude-opus-4-6_max, max, 27 August 2026
  27. 27. DeepSeek V4 Pro: 47.0% (± 1.6) — deepseek-v4-pro_max, max, 27 August 2026
  28. 28. Claude Sonnet 5.5: 46.5% (± 1.6) — claude-sonnet-5-5_max, max, 29 September 2026
  29. 29. GPT-5.4 Pro: 46.3% (± 1.6) — gpt-5.4-pro-2026-03-05_xhigh, xhigh, 27 August 2026
  30. 30. Qwen3.8-Max: 45.8% (± 1.6) — qwen3.8-max_xhigh, xhigh, 27 August 2026
  31. 31. Claude Opus 4.5: 45.7% (± 1.6) — claude-opus-4-5-20251101_32K, 32K, 27 August 2026
  32. 32. GPT-5.4: 45.1% (± 1.6) — gpt-5.4-2026-03-05_xhigh, xhigh, 27 August 2026
  33. 33. GPT-5.6 Terra: 43.2% (± 1.6) — gpt-5.6-terra_max, max, 10 August 2026
  34. 34. DeepSeek-V4.1-Flash: 42.9% (± 1.6) — deepseek-v4.1-flash_max, max, 13 October 2026
  35. 35. GPT-6 Luna: 41.4% (± 1.6) — gpt-6-luna_max, max, 22 September 2026
  36. 36. o1: 41.1% (± 1.6) — o1-2024-12-17_high, high, 31 August 2026
  37. 37. GPT-5.6 Luna: 41.0% (± 1.6) — gpt-5.6-luna_max, max, 10 August 2026
  38. 38. GPT-5.2: 37.1% (± 1.5) — gpt-5.2-2025-12-11_xhigh, xhigh, 27 August 2026
  39. 39. Claude Sonnet 4.6: 35.5% (± 1.5) — claude-sonnet-4-6_high, high, 10 August 2026
  40. 40. Claude Sonnet 5: 33.7% (± 1.5) — claude-sonnet-5_max, max, 27 August 2026
  41. 41. GPT-4.1: 31.1% (± 1.5) — gpt-4.1-2025-04-14, 31 August 2026
  42. 42. Claude Sonnet 4.5: 30.7% (± 1.5) — claude-sonnet-4-5-20250929_59K, 59K, 27 August 2026
  43. 43. Grok 4.20: 30.2% (± 1.5) — grok-4.20-0309-reasoning, 27 August 2026
  44. 44. GPT-5.4 Mini: 29.4% (± 1.4) — gpt-5.4-mini-2026-03-17_high, high, 27 August 2026
  45. 45. GPT-4o: 26.0% (± 1.4) — gpt-4o-2024-08-06, 31 August 2026
  46. 46. Claude Haiku 5.5: 23.8% (± 1.3) — claude-haiku-5-5_max, max, 9 October 2026
  47. 47. Mistral Large 4: 22.4% (± 1.3) — mistral-large-4-0-oss_high, high, 2 November 2026
  48. 48. GPT-5 Mini: 21.6% (± 1.3) — gpt-5-mini-2025-08-07_high, high, 10 August 2026
  49. 49. o4-mini: 19.6% (± 1.3) — o4-mini-2025-04-16_low, low, 27 August 2026
  50. 50. o3-mini: 15.3% (± 1.1) — o3-mini-2025-01-31_high, high, 31 August 2026
  51. 51. Claude Haiku 4.5: 13.2% (± 1.1) — claude-haiku-4-5-20251001, 10 August 2026
  52. 52. GPT-4.1 mini: 12.7% (± 1.1) — gpt-4.1-mini-2025-04-14, 31 August 2026
  53. 53. Claude 3 Opus: 12.6% (± 1.0) — claude-3-opus-20240229, 31 August 2026
  54. 54. GPT-5.4 nano: 11.7% (± 1.0) — gpt-5.4-nano-2026-03-17_high, high, 27 August 2026
  55. 54. GPT-5 nano: 11.7% (± 1.0) — gpt-5-nano-2025-08-07_high, high, 10 August 2026
  56. 56. GPT-4o mini: 8.3% (± 0.9) — gpt-4o-mini-2024-07-18, 31 August 2026
  57. 57. GPT-4.1 nano: 6.0% (± 0.8) — gpt-4.1-nano-2025-04-14, 31 August 2026

Epoch AI: the benchmark and its data · CC BY 4.0

Other benchmarks

All of the Model Tracker