University of Basel RISE · Research and Infrastructure SupportHumanities Data Benchmark

meta-llama/llama-4-scout

Metavia OpenRouter
average±4.3 (95%)
56.2
leaderboard rankof 114
≈97
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$0.49
seconds per inputmean response time
135.5
Compare with other models Its row in the Matrix All its test runs

By task

  • Data correction

    46.4

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

  • Document understanding

    49.9

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

  • Information extraction

    67.3

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

  • Transcription

    65.8

    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 74.5±19.9 ≈91of 115 $1.21$0.006 run 80.64 T1115 · 5 Jun 2026prompt.txt
Blacklist Cards 78.0±6.5 ≈103of 114 $0.25$0.008 run 366.1 T1122 · 5 Jun 2026prompt.txt
Book Advert XML Repair 5.3±5.5 ≈105of 119 $0.15$0.008 run 8.11 T1125 · 5 Jun 2026prompt.txt
Business Letters 55.0±6.9 ≈84of 114 $0.37$0.019 run 235.68 T1116 · 5 Jun 2026prompt.txt
Company Lists 69.5±16.7 ≈84of 114 $0.47$0.007 run 96.57 T1123 · 5 Jun 2026prompt.txt
Duty Rosters 62.2±7.5 ≈87of 107 $1.20$0.025 run 48.3 T1339 · 22 Jul 2026prompt.txt
Fraktur Adverts 54.9±29.7 ≈67of 114 $0.55$0.003 run 121.24 T1119 · 5 Jun 2026prompt_optimized.txt
General Meeting Minutes 59.8±26.4 ≈85of 108 $0.41$0.004 run 44.29 T1258 · 3 Jul 2026prompt.txt
Library Cards 72.0±1.5 ≈103of 114 $0.17$0.046 run 28.52 T1120 · 5 Jun 2026prompt.txt
Magazine Pages 0.0±0.0 ≈79of 115 $0.22$0.010 run 234.16 T1127 · 5 Jun 2026prompt.txt
Medieval Manuscripts 55.0±15.4 ≈72of 114 $0.29$0.004 run 70.46 T1121 · 5 Jun 2026prompt.txt
Personnel Cards 87.5±6.1 ≈78of 114 $0.54$0.033 run 291.7 T1126 · 5 Jun 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
22 Jul 20260.10.3archived price page
5 Jun 20260.080.3archived price page

Every run

  1. T1339Duty Rosters22 Jul 202662.2$0.025
  2. T1258General Meeting Minutes3 Jul 202659.8$0.004
  3. T1127Magazine Pages5 Jun 20260.0$0.010
  4. T1126Personnel Cards5 Jun 202687.5$0.033
  5. T1125Book Advert XML Repair5 Jun 20265.3$0.008
  6. T1124Company Lists5 Jun 202647.7$0.007
  7. T1123Company Lists5 Jun 202669.5$0.007
  8. T1122Blacklist Cards5 Jun 202678.0$0.008
  9. T1121Medieval Manuscripts5 Jun 202655.0$0.004
  10. T1120Library Cards5 Jun 202672.0$0.046
  11. T1119Fraktur Adverts5 Jun 202654.9$0.003
  12. T1118Business Letters5 Jun 2026 · 32 signed letters59.0$0.013
  13. T1117Business Letters5 Jun 2026 · 25 unsigned letters51.0$0.007
  14. T1116Business Letters5 Jun 2026 · all 57 letters55.0$0.019
  15. T1115Bibliographic Data5 Jun 202674.5$0.006