University of Basel RISE · Research and Infrastructure SupportHumanities Data Benchmark

nvidia/nemotron-3.5-super-vl

NVIDIAvia OpenRouter
average±3.6 (95%)
74.0
leaderboard rankof 114
≈44
benchmarksof 12
12
per 1,000 documentsmean over its benchmarks
$4.04
seconds per inputmean response time
39.9
Compare with other models Its row in the Matrix All its test runs

By task

  • Data correction

    67.8

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

  • Document understanding

    80.0

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

  • Information extraction

    81.2

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

  • Transcription

    70.1

    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 Dataother prompt 79.5±14.4 ≈91of 115 $3.52$0.018 run 26.96 T2012 · 24 Oct 2026prompt_v2.txt
Blacklist Cards 92.1±2.6 ≈51of 114 $1.28$0.042 run 8.6 T2019 · 24 Oct 2026prompt.txt
Book Advert XML Repairother prompt 41.6±11.5 ≈16of 119 $4.01$0.20 run 111.79 T2021 · 24 Oct 2026prompt_v2.txt
Business Letters 60.0±6.0 ≈24of 114 $2.54$0.14 run 18.59 T2013 · 24 Oct 2026prompt.txt
Company Listsother prompt 89.5±12.4 ≈6of 114 $2.84$0.043 run 29.02 T2020 · 24 Oct 2026prompt_v2.txt
Duty Rosters 78.2±4.7 ≈46of 107 $9.24$0.19 run 86.0 T2025 · 24 Oct 2026prompt.txt
Fraktur Adverts 51.0±30.7 ≈67of 114 $12$0.059 run 71.94 T2016 · 24 Oct 2026prompt_optimized.txt
General Meeting Minutes 82.2±9.4 ≈25of 108 $3.08$0.028 run 50.52 T2024 · 24 Oct 2026prompt.txt
Library Cardsother prompt 87.0±1.5 ≈37of 114 $1.51$0.40 run 9.9 T2017 · 24 Oct 2026prompt_year-string.txt
Magazine Pagesother prompt 67.6±13.3 ≈34of 115 $1.75$0.081 run 15.15 T2023 · 24 Oct 2026prompt_explicit-json.txt
Medieval Manuscripts 65.3±10.5 ≈72of 114 $3.50$0.042 run 27.44 T2018 · 24 Oct 2026prompt.txt
Personnel Cards 94.0±1.6 ≈58of 114 $3.38$0.21 run 22.91 T2022 · 24 Oct 2026prompt.txt

Price

FromInput, $ per million tokensOutput, $ per million tokensSource
24 Oct 20260.150.5archived price page

Every run

  1. T2025Duty Rosters24 Oct 202678.2$0.19
  2. T2024General Meeting Minutes24 Oct 202682.2$0.028
  3. T2023Magazine Pages24 Oct 202667.6$0.081
  4. T2022Personnel Cards24 Oct 202694.0$0.21
  5. T2021Book Advert XML Repair24 Oct 202641.6$0.20
  6. T2020Company Lists24 Oct 202689.5$0.043
  7. T2019Blacklist Cards24 Oct 202692.1$0.042
  8. T2018Medieval Manuscripts24 Oct 202665.3$0.042
  9. T2017Library Cards24 Oct 202687.0$0.40
  10. T2016Fraktur Adverts24 Oct 202651.0$0.059
  11. T2015Business Letters24 Oct 2026 · 32 signed letters64.0$0.11
  12. T2014Business Letters24 Oct 2026 · 25 unsigned letters62.0$0.063
  13. T2013Business Letters24 Oct 2026 · all 57 letters60.0$0.14
  14. T2012Bibliographic Data24 Oct 202679.5$0.018