RISE Humanities Data Benchmark, 0.6.1

Search Test Runs

 

A test run is a single execution of a benchmark test using a defined model configuration.
Each run represents how a particular large language model (LLM) — such as GPT-4, Claude-3, or Gemini — performed on a given task at a specific time, with specific settings.

A test run includes:

  • Prompt and role definition – what the model was asked to do and from what perspective (e.g. “as a historian”).
  • Model configuration – provider, model version, temperature, and other generation parameters.
  • Results – the model’s actual response and its evaluation (scores such as F1 or accuracy).
  • Usage and cost data – token counts and calculated API costs.
  • Metadata – information like the test date, benchmark name, and person who executed it.

Together, test runs make it possible to compare models, providers, and configurations across benchmarks in a transparent and reproducible way.

Loading module...

Search Results

Your search for Benchmark 'medieval_manuscripts__true' with Search Hidden 'False' returned 177 results, showing page 18 of 18.
Result 171 of 177

Test T0271 at 2025-10-24

manuscript transcription handwritten 15 de prose

genai gemini-2.5-flash · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

{
  "folios": [
    {
      "folio": "3",
      "text": "Vnd ein pferit Die mir vnd
minen knechten vber Rulfand
Sen So was mencky kein weg
Denne den wir machtend
vnd ziehend Sie knecht die
vnd vil in vntz an den and
vnd die pferit vntz an die
bettel vnd was ze mol ein grosse
nebel do wir kaum gesahend
vnd als mit grosser arbeit kame
wir ze mittem tag zu sant
kinroffel vff den berg Do
do sach ich die buecher do gar
vil herren wopen zu stonde
die in four edlen geley hand
do seind mincs vattero pflegen
wopen ach in dem einem",
      "addition1": "",
      "addition2": "",
      "addition3": ""
    }
  ]
}

Scoring

Fuzzy score69.0%
CER ↓34.0%

Speed

Model time, all inputs185.4 s
Mean per input15.45 s
Slowest input47.58 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens9.1K
Output tokens3.4K
Total tokens12.5K
Input cost$0.0027
Output cost$0.0084
Total cost$0.066
Reasoning cost usd$0.0548
Total reasoning tokens21.9K

Priced from 2025-10-01 · Input $/M $0.30 · Output $/M $2.50

Result 172 of 177

Test T0273 at 2025-10-24

manuscript transcription handwritten 15 de prose

openai gpt-4.1 · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

Scoring

Fuzzy score64.8%
CER ↓39.8%

Speed

Model time, all inputs94.0 s
Mean per input7.84 s
Slowest input10.92 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens17.1K
Output tokens2.4K
Total tokens19.5K
Input cost$0.0343
Output cost$0.0192
Total cost$0.0535

Priced from 2025-10-01 · Input $/M $2.00 · Output $/M $8.00

Result 173 of 177

Test T0285 at 2025-10-24

manuscript transcription handwritten 15 de prose

genai gemini-2.5-flash-lite · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

{
  "folios": [
    {
      "folio": "3",
      "text": "vnd ein pferit die mir vnd
minen enertgen vber Rulpend
Son So wab mencky ein weg
Semme Den wir machtend
und viehend Sie Enertgt Snd
vnd vil in vntz an Den vnd
ond Sie pferit vntz an Die
petel vnd wab ze mol ein groffe
nebel So wir kum gepaussend
vnd alſo mit groffer arbeit kame
wir ze mittem tag zu Sant
Einroffel vff Say berg So
Do fach ich die Bücher So jar
vil Gerren wopay zu pone
Die w ſuir Solgen gelben band
So ſaint minek vatter Pligen
wopay ach in Sem emeng",
      "addition1": "",
      "addition2": "",
      "addition3": ""
    }
  ]
}

Scoring

Fuzzy score58.2%
CER ↓43.0%

Speed

Model time, all inputs43.3 s
Mean per input3.61 s
Slowest input5.24 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens9.1K
Output tokens2.9K
Total tokens12K
Input cost$0.0009
Output cost$0.0012
Total cost$0.0021

Priced from 2025-10-01 · Input $/M $0.10 · Output $/M $0.40

Result 174 of 177

Test T0280 at 2025-10-24

manuscript transcription handwritten 15 de prose

openai gpt-5-mini · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

Scoring

Fuzzy score71.4%
CER ↓33.4%

Speed

Model time, all inputs507.8 s
Mean per input42.32 s
Slowest input77.39 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens18.1K
Output tokens28.6K
Total tokens46.7K
Input cost$0.0045
Output cost$0.0572
Total cost$0.0618

Priced from 2025-10-01 · Input $/M $0.25 · Output $/M $2.00

Result 175 of 177

Test T0277 at 2025-10-24

manuscript transcription handwritten 15 de prose

openai gpt-4.1-mini · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

Scoring

Fuzzy score79.8%
CER ↓23.1%

Speed

Model time, all inputs52.7 s
Mean per input4.39 s
Slowest input8.49 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens21.7K
Output tokens2.4K
Total tokens24K
Input cost$0.0087
Output cost$0.0038
Total cost$0.0125

Priced from 2025-10-01 · Input $/M $0.40 · Output $/M $1.60

Result 176 of 177

Test T0272 at 2025-10-24

manuscript transcription handwritten 15 de prose

genai gemini-2.5-pro · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

{
  "folios": [
    {
      "folio": "3",
      "text": "vnd ein pferit Die mir vnd
minen knechten vber Rulpend
Sen So was menen kein weg
Sonne Den wir machtend
vnd viehend Sie knecht vnd
vnd vil in vntz an den anß
vnd Sie pferit vntz an die
petel vnd was ze mol ein grosse
nebel dz wir kum gepaßend
vnd alß mit grosser arbeit kome
wir ze mittem tag zu sant
kristoffel vff den berg So
do sach ich die burger So gar
vil gerren wopen zu pont
die ir fúr Solrin geben hand
So seind mines vatterß flugen
wopen ouch in dem einen",
      "addition1": null,
      "addition2": null,
      "addition3": null
    }
  ]
}

Scoring

Fuzzy score68.4%
CER ↓33.9%

Speed

Model time, all inputs276.6 s
Mean per input23.05 s
Slowest input47.28 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens9.1K
Output tokens2.8K
Total tokens11.9K
Input cost$0.0114
Output cost$0.0277
Total cost$0.3198
Reasoning cost usd$0.2806
Total reasoning tokens28.1K

Priced from 2025-10-01 · Input $/M $1.25 · Output $/M $10.00

Result 177 of 177

Test T0293 at 2025-10-24

manuscript transcription handwritten 15 de prose

anthropic claude-sonnet-4-5-20250929 · temp 0.0 · dataclass Document

Response vs ground truth

One model response from this run. The per-input comparison needs JavaScript.

Scoring

Fuzzy score55.0%
CER ↓48.9%

Speed

Model time, all inputs161.1 s
Mean per input13.43 s
Slowest input23.35 s
Inputs timed12

Summed model time, not elapsed: requests run in parallel and the worker count is not recorded, so the run finished sooner than this.

Costs / Pricing

Input tokens27.1K
Output tokens3.8K
Total tokens30.9K
Input cost$0.0814
Output cost$0.0564
Total cost$0.1378

Priced from 2025-10-01 · Input $/M $3.00 · Output $/M $15.00