gpt-5

OpenAI
average
71.4
leaderboard rankof 114
≈61
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$42
seconds per inputmean response time
85.7

By task

  • Data correction

    55.9

    Best of 114 models with all its benchmarks: claude-opus-5-5, 82.7.

  • Document understanding

    84.6

    Best of 107 models with all its benchmarks: gpt-6-astra, 97.0.

  • Information extraction

    86.6

    Best of 107 models with all its benchmarks: grok-4.6, 90.8.

  • Transcription

    51.2

    Best of 114 models with all its benchmarks: gemini-4-argon, 96.2.

By benchmark

BenchmarkScoreThe best current run on each benchmark, on the benchmark's metric scaled to 0–100. ± is the 95% sampling margin, from resampling the run's inputs.RankRank among all current models on that benchmark; ≈ for models that cannot be told apart within sampling error.Per 1,000 docss / inputBest run
Bibliographic Data 95.7±4.0 ≈50of 115 $65$0.33 run 147.04 T0129 · 25 Jan 2026prompt.txt
Blacklist Cards 93.8±2.4 ≈45of 114 $14$0.45 run 34.79 T0309 · 26 Jan 2026prompt.txt
Book Advert XML Repair 17.9±7.2 ≈64of 119 $53$2.64 run 99.4 T0451 · 8 Dec 2025prompt.txt
Business Lettersolder run 71.0 ≈7of 114 $23$1.30 run 51.27 T0108 · 30 Sep 2025prompt.txt
Company Lists 95.7±2.7 ≈1of 114 $42$0.63 run 111.73 T0347 · 25 Jan 2026prompt.txt
Duty Rosters 81.4±2.9 ≈30of 107 $62$1.30 run 98.53 T1298 · 16 Sep 2026prompt.txt
Fraktur Advertsolder run 13.6 103of 114 $41$0.20 run 98.79 T0120 · 30 Sep 2025prompt_optimized.txt
General Meeting Minutes 78.9±14.5 ≈25of 108 $50$0.45 run 67.5 T1249 · 16 Sep 2026prompt.txt
Library Cardsolder run 89.6 ≈15of 114 –– run 48.46 T0165 · 2 Sep 2025prompt.txt
Magazine Pages 78.7±9.7 ≈18of 115 $24$1.10 run 45.96 T0716 · 21 Mar 2026prompt.txt
Medieval Manuscripts 46.2±18.4 ≈95of 114 $54$0.65 run 171.35 T0279 · 26 Jan 2026prompt.txt
Personnel Cards 93.9±2.5 ≈58of 114 $39$2.36 run 53.76 T0610 · 9 Feb 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
0 Sep 20251.2510.0archived price page

Every run

  1. T1298Duty Rosters16 Sep 202681.4$1.30
  2. T1249General Meeting Minutes16 Sep 202678.9$0.45
  3. T0716Magazine Pages16 Sep 202672.4$1.16
  4. T0610Personnel Cards16 Sep 202693.1$2.38
  5. T1298Duty Rosters24 Jul 202680.7$1.21
  6. T1249General Meeting Minutes3 Jul 202677.0$0.52
  7. T0716Magazine Pages21 Mar 202678.7$1.10
  8. T0109Business Letters17 Feb 2026 · 25 unsigned letters92.0$0.43
  9. T0109Business Letters10 Feb 2026 · 25 unsigned letters87.0$0.47
  10. T0610Personnel Cards9 Feb 202693.9$2.36
  11. T0309Blacklist Cards26 Jan 202693.8$0.45
  12. T0279Medieval Manuscripts26 Jan 202646.2$0.65
  13. T0165Library Cards26 Jan 202659.7$4.50
  14. T0110Business Letters26 Jan 2026 · 32 signed letters66.0$0.72
  15. T0108Business Letters26 Jan 2026 · all 57 letters69.0$1.12
  16. T0451Book Advert XML Repair25 Jan 202617.0$2.75
  17. T0348Company Lists25 Jan 202673.4$0.56
  18. T0347Company Lists25 Jan 202695.7$0.63
  19. T0129Bibliographic Data25 Jan 202695.7$0.33
  20. T0120Fraktur Adverts25 Jan 202612.8$0.24
  21. T0451Book Advert XML Repair8 Dec 202517.9$2.64
  22. T0348Company Lists28 Oct 202567.1$0.53
  23. T0347Company Lists28 Oct 202593.4$0.72
  24. T0309Blacklist Cards24 Oct 202591.9$0.51
  25. T0279Medieval Manuscripts24 Oct 202541.9$0.53
  26. T0165Library Cards1 Oct 202586.5$7.12
  27. T0129Bibliographic Data1 Oct 202594.0$0.34
  28. T0129Bibliographic Data30 Sep 202595.5$0.34
  29. T0120Fraktur Adverts30 Sep 202513.6$0.20
  30. T0110Business Letters30 Sep 2025 · 32 signed letters70.0$0.77
  31. T0109Business Letters30 Sep 2025 · 25 unsigned letters77.0$0.48
  32. T0108Business Letters30 Sep 2025 · all 57 letters71.0$1.30
  33. T0120Fraktur Adverts24 Sep 20250.0–
  34. T0165Library Cards2 Sep 202589.6–
  35. T0129Bibliographic Data14 Aug 202592.7–
  36. T0120Fraktur Adverts13 Aug 20255.0–
  37. T0110Business Letters13 Aug 2025 · 32 signed letters69.0–
  38. T0109Business Letters13 Aug 2025 · 25 unsigned letters93.0–
  39. T0108Business Letters13 Aug 2025 · all 57 letters70.0–