University of Basel RISE · Research and Infrastructure SupportHumanities Data Benchmark

claude-sonnet-5-5

Anthropic
average±1.8 (95%)
86.9
leaderboard rankof 114
≈8
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$21
seconds per inputmean response time
9.1
Compare with other models Its row in the Matrix All its test runs

By task

  • Data correction

    76.5

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

  • Document understanding

    91.8

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

  • Information extraction

    88.4

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

  • Transcription

    95.2

    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 98.3±0.8 ≈8of 115 $28$0.14 run 9.36 T1870 · 29 Sep 2026prompt.txt
Blacklist Cards 96.7±2.2 ≈8of 114 $8.58$0.28 run 2.64 T1877 · 28 Sep 2026prompt.txt
Book Advert XML Repair 55.4±9.7 ≈1of 119 $19$0.95 run 11.21 T1880 · 29 Sep 2026prompt.txt
Business Letters 60.0±6.8 ≈24of 114 $15$0.88 run 2.94 T1871 · 28 Sep 2026prompt.txt
Company Lists 84.8±12.3 ≈29of 114 $19$0.29 run 6.45 T1878 · 28 Sep 2026prompt.txt
Duty Rosters 95.3±1.4 ≈5of 107 $41$0.86 run 25.87 T1884 · 29 Sep 2026prompt.txt
Fraktur Adverts 99.4±0.1 ≈1of 114 $53$0.26 run 26.4 T1874 · 28 Sep 2026prompt_optimized.txt
General Meeting Minutes 92.2±5.2 ≈1of 108 $20$0.18 run 6.65 T1883 · 29 Sep 2026prompt.txt
Library Cardsother prompt 91.7±1.4 ≈1of 114 $5.04$1.32 run 2.9 T1875 · 28 Sep 2026prompt_year-string.txt
Magazine Pagesother prompt 82.5±10.5 ≈18of 115 $11$0.48 run 3.32 T1882 · 29 Sep 2026prompt_explicit-json.txt
Medieval Manuscripts 88.7±6.4 ≈1of 114 $12$0.15 run 4.73 T1876 · 28 Sep 2026prompt.txt
Personnel Cards 97.7±1.0 ≈1of 114 $25$1.50 run 6.6 T1881 · 29 Sep 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
28 Sep 20262.010.0archived price page

Every run

  1. T1884Duty Rosters29 Sep 202695.3$0.86
  2. T1883General Meeting Minutes29 Sep 202692.2$0.18
  3. T1882Magazine Pages29 Sep 202682.5$0.48
  4. T1881Personnel Cards29 Sep 202697.7$1.50
  5. T1880Book Advert XML Repair29 Sep 202655.4$0.95
  6. T1879Company Lists29 Sep 202681.5$0.29
  7. T1878Company Lists29 Sep 202683.4$0.31
  8. T1877Blacklist Cards29 Sep 202696.5$0.37
  9. T1876Medieval Manuscripts29 Sep 202688.0$0.16
  10. T1875Library Cards29 Sep 202691.2$2.23
  11. T1874Fraktur Adverts29 Sep 202699.4$0.20
  12. T1873Business Letters29 Sep 2026 · 32 signed letters64.0$0.61
  13. T1872Business Letters29 Sep 2026 · 25 unsigned letters57.0$0.36
  14. T1871Business Letters29 Sep 2026 · all 57 letters60.0$0.97
  15. T1870Bibliographic Data29 Sep 202698.3$0.14
  16. T1884Duty Rosters28 Sep 20261.4$0.88
  17. T1883General Meeting Minutes28 Sep 202625.6$0.26
  18. T1882Magazine Pages28 Sep 202664.4$0.49
  19. T1881Personnel Cards28 Sep 202697.5$1.48
  20. T1880Book Advert XML Repair28 Sep 202646.1$1.25
  21. T1879Company Lists28 Sep 20261.9$0.34
  22. T1878Company Lists28 Sep 202684.8$0.29
  23. T1877Blacklist Cards28 Sep 202696.7$0.28
  24. T1876Medieval Manuscripts28 Sep 202688.7$0.15
  25. T1875Library Cards28 Sep 202691.7$1.32
  26. T1874Fraktur Adverts28 Sep 202699.4$0.26
  27. T1873Business Letters28 Sep 2026 · 32 signed letters64.0$0.56
  28. T1872Business Letters28 Sep 2026 · 25 unsigned letters58.0$0.32
  29. T1871Business Letters28 Sep 2026 · all 57 letters60.0$0.88
  30. T1870Bibliographic Data28 Sep 202643.8$0.16