University of Basel RISE · Research and Infrastructure SupportHumanities Data Benchmark

meta/muse-spark-1.2

Metavia OpenRouter
average±1.6 (95%)
84.8
leaderboard rankof 114
≈8
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$20
seconds per inputmean response time
50.9
Compare with other models Its row in the Matrix All its test runs

By task

  • Data correction

    66.6

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

  • Document understanding

    90.5

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

  • Information extraction

    87.7

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

  • Transcription

    93.6

    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±5.9 ≈50of 115 $24$0.12 run 256.83 T1615 · 18 Aug 2026prompt.txt
Blacklist Cards 96.2±2.3 ≈14of 114 $9.48$0.31 run 5.6 T1622 · 17 Sep 2026prompt.txt
Book Advert XML Repair 35.2±9.4 ≈41of 119 $16$0.82 run 11.81 T1625 · 17 Sep 2026prompt.txt
Business Letters 65.0±7.3 ≈21of 114 $19$1.06 run 24.39 T1616 · 18 Aug 2026prompt.txt
Company Lists 90.1±5.5 ≈6of 114 $17$0.26 run 27.42 T1623 · 18 Aug 2026prompt.txt
Duty Rosters 85.6±3.4 ≈18of 107 $34$0.71 run 94.36 T1629 · 18 Aug 2026prompt.txt
Fraktur Adverts 97.7±1.0 ≈15of 114 $32$0.16 run 21.75 T1619 · 17 Sep 2026prompt_optimized.txt
General Meeting Minutes 91.8±5.5 ≈1of 108 $18$0.16 run 53.79 T1628 · 18 Aug 2026prompt.txt
Library Cards 89.2±1.5 ≈15of 114 $14$3.72 run 39.86 T1620 · 17 Sep 2026prompt.txt
Magazine Pages 87.8±7.8 ≈18of 115 $16$0.71 run 52.06 T1627 · 18 Aug 2026prompt.txt
Medieval Manuscripts 85.1±7.2 ≈1of 114 $18$0.22 run 12.38 T1621 · 17 Sep 2026prompt.txt
Personnel Cards 98.1±0.9 ≈1of 114 $17$1.05 run 10.02 T1626 · 17 Sep 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
18 Aug 20261.254.25archived price page

Every run

  1. T1629Duty Rosters17 Sep 202685.0$0.71
  2. T1628General Meeting Minutes17 Sep 202684.6$0.16
  3. T1627Magazine Pages17 Sep 202686.8$0.47
  4. T1626Personnel Cards17 Sep 202698.1$1.05
  5. T1625Book Advert XML Repair17 Sep 202635.2$0.82
  6. T1624Company Lists17 Sep 202669.2$0.21
  7. T1623Company Lists17 Sep 202681.6$0.32
  8. T1622Blacklist Cards17 Sep 202696.2$0.31
  9. T1621Medieval Manuscripts17 Sep 202685.1$0.22
  10. T1620Library Cards17 Sep 202689.2$3.72
  11. T1619Fraktur Adverts17 Sep 202697.7$0.16
  12. T1618Business Letters17 Sep 2026 · 32 signed letters67.0$0.69
  13. T1617Business Letters17 Sep 2026 · 25 unsigned letters68.0$0.41
  14. T1616Business Letters17 Sep 2026 · all 57 letters62.0$1.06
  15. T1615Bibliographic Data17 Sep 202679.2$0.12
  16. T1629Duty Rosters18 Aug 202685.6$0.71
  17. T1628General Meeting Minutes18 Aug 202691.8$0.16
  18. T1627Magazine Pages18 Aug 202687.8$0.71
  19. T1626Personnel Cards18 Aug 202698.0$0.95
  20. T1625Book Advert XML Repair18 Aug 202630.6$0.82
  21. T1624Company Lists18 Aug 202640.5$0.20
  22. T1623Company Lists18 Aug 202690.1$0.26
  23. T1622Blacklist Cards18 Aug 202695.6$0.29
  24. T1621Medieval Manuscripts18 Aug 202684.1$0.18
  25. T1620Library Cards18 Aug 202689.1$4.77
  26. T1619Fraktur Adverts18 Aug 202688.2$0.14
  27. T1618Business Letters18 Aug 2026 · 32 signed letters67.0$0.63
  28. T1617Business Letters18 Aug 2026 · 25 unsigned letters64.0$0.40
  29. T1616Business Letters18 Aug 2026 · all 57 letters65.0$1.06
  30. T1615Bibliographic Data18 Aug 202695.7$0.12