University of Basel RISE · Research and Infrastructure SupportHumanities Data Benchmark

gemini-3.1-pro-preview

Google
average±1.9 (95%)
78.8
leaderboard rankof 114
≈37
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$50
seconds per inputmean response time
29.2
Compare with other models Its row in the Matrix All its test runs

By task

  • Data correction

    74.2

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

  • Document understanding

    62.5

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

  • Information extraction

    87.5

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

  • Transcription

    94.0

    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 88.3±2.6 ≈74of 115 $33$0.17 run 18.17 T0681 · 16 Sep 2026prompt.txt
Blacklist Cards 95.4±2.6 ≈14of 114 $4.62$0.15 run 3.77 T0688 · 16 Sep 2026prompt.txt
Book Advert XML Repair 51.4±10.1 ≈16of 119 $86$4.31 run 54.7 T0691 · 16 Mar 2026prompt.txt
Business Letters 68.0±7.8 ≈10of 114 $35$2.02 run 22.04 T0682 · 16 Mar 2026prompt.txt
Company Lists 87.7±13.1 ≈29of 114 $17$0.26 run 11.43 T0689 · 16 Mar 2026prompt.txt
Duty Rosters 90.3±2.8 ≈10of 107 $281$5.90 run 151.0 T1311 · 24 Jul 2026prompt.txt
Fraktur Adverts 99.1±0.4 ≈7of 114 $64$0.32 run 37.7 T0685 · 16 Sep 2026prompt_optimized.txt
General Meeting Minutes 94.4±3.6 ≈1of 108 $23$0.21 run 13.55 T1233 · 16 Sep 2026prompt.txt
Library Cards 88.3±1.6 ≈15of 114 $25$6.64 run 16.92 T0686 · 16 Mar 2026prompt.txt
Magazine Pages 0.0±0.0 ≈79of 115 $3.16$0.15 run 4.45 T0711 · 21 Mar 2026prompt.txt
Medieval Manuscripts 86.0±12.4 ≈1of 114 $5.95$0.071 run 5.06 T0687 · 16 Sep 2026prompt.txt
Personnel Cards 97.0±1.1 ≈30of 114 $19$1.15 run 11.35 T0692 · 16 Mar 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
16 Mar 20262.012.0archived price page

Every run

  1. T1311Duty Rosters16 Sep 202688.6$3.91
  2. T1233General Meeting Minutes16 Sep 202694.4$0.21
  3. T0711Magazine Pages16 Sep 20260.0$0.15
  4. T0692Personnel Cards16 Sep 202696.0$1.05
  5. T0691Book Advert XML Repair16 Sep 202650.2$3.48
  6. T0690Company Lists16 Sep 202674.1$0.22
  7. T0689Company Lists16 Sep 202684.2$0.29
  8. T0688Blacklist Cards16 Sep 202695.4$0.15
  9. T0687Medieval Manuscripts16 Sep 202686.0$0.071
  10. T0686Library Cards16 Sep 202688.0$3.77
  11. T0685Fraktur Adverts16 Sep 202699.1$0.32
  12. T0684Business Letters16 Sep 2026 · 32 signed letters66.0$0.74
  13. T0683Business Letters16 Sep 2026 · 25 unsigned letters72.0$0.19
  14. T0682Business Letters16 Sep 2026 · all 57 letters67.0$0.55
  15. T0681Bibliographic Data16 Sep 202688.3$0.17
  16. T1311Duty Rosters24 Jul 202690.3$5.90
  17. T1233General Meeting Minutes3 Jul 202689.4$0.13
  18. T0711Magazine Pages21 Mar 20260.0$0.15
  19. T0692Personnel Cards16 Mar 202697.0$1.15
  20. T0691Book Advert XML Repair16 Mar 202651.4$4.31
  21. T0690Company Lists16 Mar 202660.7$1.76
  22. T0689Company Lists16 Mar 202687.7$0.26
  23. T0688Blacklist Cards16 Mar 202693.4$0.94
  24. T0687Medieval Manuscripts16 Mar 2026––
  25. T0686Library Cards16 Mar 202688.3$6.64
  26. T0685Fraktur Adverts16 Mar 202699.0$0.21
  27. T0684Business Letters16 Mar 2026 · 32 signed letters67.0$1.81
  28. T0683Business Letters16 Mar 2026 · 25 unsigned letters78.0$0.20
  29. T0682Business Letters16 Mar 2026 · all 57 letters68.0$2.02
  30. T0681Bibliographic Data16 Mar 202688.0$0.27