University of Basel RISE · Research and Infrastructure SupportHumanities Data Benchmark

claude-sonnet-4-6

Anthropic
average±1.9 (95%)
75.2
leaderboard rankof 114
≈44
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$22
seconds per inputmean response time
17.5
Compare with other models Its row in the Matrix All its test runs

By task

  • Data correction

    68.0

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

  • Document understanding

    59.5

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

  • Information extraction

    83.6

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

  • Transcription

    91.4

    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 97.6±2.0 ≈8of 115 $37$0.19 run 25.89 T0640 · 16 Sep 2026prompt.txt
Blacklist Cards 97.4±2.3 ≈1of 114 $11$0.38 run 5.35 T0647 · 16 Mar 2026prompt.txt
Book Advert XML Repair 39.0±8.4 ≈41of 119 $18$0.92 run 12.46 T0650 · 16 Mar 2026prompt.txt
Business Letters 59.0±6.1 ≈24of 114 $13$0.75 run 4.38 T0641 · 16 Mar 2026prompt.txt
Company Lists 81.9±13.5 ≈58of 114 $24$0.36 run 16.39 T0648 · 16 Sep 2026prompt.txt
Duty Rosters 78.2±3.9 ≈46of 107 $34$0.72 run 39.51 T1309 · 24 Jul 2026prompt.txt
Fraktur Adverts 98.3±1.6 ≈9of 114 $44$0.22 run 56.65 T0644 · 16 Mar 2026prompt_optimized.txt
General Meeting Minutes 84.8±8.5 ≈25of 108 $24$0.22 run 17.39 T1217 · 17 Sep 2026prompt.txt
Library Cards 86.5±1.4 ≈41of 114 $11$2.98 run 5.18 T0645 · 16 Sep 2026prompt.txt
Magazine Pages 3.3±4.7 ≈79of 115 $8.45$0.39 run 4.47 T0712 · 21 Mar 2026prompt.txt
Medieval Manuscripts 78.9±9.6 ≈27of 114 $13$0.16 run 11.4 T0646 · 16 Mar 2026prompt.txt
Personnel Cards 97.0±1.3 ≈30of 114 $27$1.66 run 11.4 T0651 · 16 Mar 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
16 Mar 20263.015.0archived price page

Every run

  1. T1309Duty Rosters17 Sep 202678.2$0.73
  2. T1217General Meeting Minutes17 Sep 202684.8$0.22
  3. T0643Business Letters17 Sep 2026 · 32 signed letters63.0$0.46
  4. T0642Business Letters17 Sep 2026 · 25 unsigned letters55.0$0.29
  5. T0641Business Letters17 Sep 2026 · all 57 letters59.0$0.75
  6. T0712Magazine Pages16 Sep 20261.1$0.39
  7. T0651Personnel Cards16 Sep 202696.5$1.66
  8. T0650Book Advert XML Repair16 Sep 202635.7$0.91
  9. T0649Company Lists16 Sep 202662.4$0.32
  10. T0648Company Lists16 Sep 202681.9$0.36
  11. T0647Blacklist Cards16 Sep 202697.2$0.38
  12. T0646Medieval Manuscripts16 Sep 202678.8$0.16
  13. T0645Library Cards16 Sep 202686.5$2.98
  14. T0644Fraktur Adverts16 Sep 202698.1$0.22
  15. T0643Business Letters16 Sep 2026 · 32 signed letters63.0$0.46
  16. T0642Business Letters16 Sep 2026 · 25 unsigned letters55.0$0.29
  17. T0641Business Letters16 Sep 2026 · all 57 letters59.0$0.75
  18. T0640Bibliographic Data16 Sep 202697.6$0.19
  19. T1309Duty Rosters24 Jul 202678.2$0.72
  20. T1217General Meeting Minutes3 Jul 202683.5$0.22
  21. T0712Magazine Pages21 Mar 20263.3$0.39
  22. T0651Personnel Cards16 Mar 202697.0$1.66
  23. T0650Book Advert XML Repair16 Mar 202639.0$0.92
  24. T0649Company Lists16 Mar 202667.2$0.32
  25. T0648Company Lists16 Mar 202676.9$0.36
  26. T0647Blacklist Cards16 Mar 202697.4$0.38
  27. T0646Medieval Manuscripts16 Mar 202678.9$0.16
  28. T0645Library Cards16 Mar 202686.4$2.98
  29. T0644Fraktur Adverts16 Mar 202698.3$0.22
  30. T0643Business Letters16 Mar 2026 · 32 signed letters61.0$0.46
  31. T0642Business Letters16 Mar 2026 · 25 unsigned letters55.0$0.29
  32. T0641Business Letters16 Mar 2026 · all 57 letters59.0$0.75
  33. T0640Bibliographic Data16 Mar 202697.3$0.19