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 14 of 18.
Result 131 of 177

Test T0494 at 2026-03-23

manuscript transcription handwritten 15 de prose

openai gpt-5.2-2025-12-11 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score94.5%
CER ↓5.6%

{"folios":[{"folio":"3","text":"Vnd ein pferit die mir vnd
minen knechten vber hulffend
Den So was meinen kein weg
danne den wir machten
vnd vielend die knecht dick
vnd vil in vntz an den aꝛs
vnd die pferit vntz an die
sitel vnd was ze mol ein grosser
nebel Es wir kum gesachend
vnd als mit grosser arbeit kame
wir ze mitten tag zů Sant
kꝛistoffel vff den berg So
do sach ich die bücher So gar
vil herren wopen jn ston
Die in für S. kꝛistoffel glichend hand
So stund mines vatters schilgen
wopen och in dem einich","addition1":"","addition2":"","addition3":""}]}

Scoring

Fuzzy score70.1%
CER ↓31.9%

Speed

Model time, all inputs99.6 s
Mean per input8.30 s
Slowest input15.36 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 tokens28.9K
Output tokens2.9K
Total tokens31.8K
Input cost$0.0506
Output cost$0.0409
Total cost$0.0915

Priced from 2026-03-16 · Input $/M $1.75 · Output $/M $14.00

Result 132 of 177

Test T0699 at 2026-03-17

manuscript transcription handwritten 15 de prose

openai gpt-5.3-codex · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score94.5%
CER ↓5.1%

{
  "folios": [
    {
      "folio": "3",
      "text": "Vnd ein pferit die mir vnd
minen knechten vber hulffend
den So was nieman kein weg
Denne den wir machtenß
vnd vielend die knechte dick
vnd vil in vntz an den arß
vnd die pferit vntz an die
setel vnd was ze mol ein grosser
nebel Ez wir kaum geſachend
vnd alß mit grosser arbeit kame
wir ze mitten tag zů Sant
kristoffel vff den berg do
do hess ich die bůcher So gar
vil herren wapen jn stonnd
Die ir für delfin gelien hand
So ſtůnd mineß vatters klůgen
wapen och in dem crentz",
      "addition1": "",
      "addition2": "",
      "addition3": ""
    }
  ]
}

Scoring

Fuzzy score74.6%
CER ↓28.5%

Speed

Model time, all inputs100.8 s
Mean per input8.40 s
Slowest input13.87 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 tokens29.8K
Output tokens2.9K
Total tokens32.7K
Input cost$0.0521
Output cost$0.0401
Total cost$0.0923

Priced from 2026-03-16 · Input $/M $1.75 · Output $/M $14.00

Result 133 of 177

Test T0634 at 2026-03-16

manuscript transcription handwritten 15 de prose

anthropic claude-opus-4-6 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score97.4%
CER ↓1.3%

{"folios":[{"folio":"3","text":"vnd ein pferit die mir vnd
minen knechten vber hulffend
den do was menen kein weg
denne den wir machtend
vnd viehend die knecht dick
vnd vil in vntz an den arß
vnd die pferit vntz an die
settel vnd was ze mol ein grosseꝛ
nebel dz wir kum gesachend
vnd also mit grosser arbeit kame
wir ze mittem tag zuͦ sant
kristoffel vff den berg do
do sach ich die buͦcher do gar
vil herren wopen jn stond
die ir stür do hin geben hand
do stuͦnd mines vatters seligen
wopen och in dem einen","addition1":null,"addition2":null,"addition3":null}]}

Scoring

Fuzzy score80.7%
CER ↓20.2%

Speed

Model time, all inputs159.5 s
Mean per input13.29 s
Slowest input22.23 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 tokens35.9K
Output tokens3.3K
Total tokens39.2K
Input cost$0.1795
Output cost$0.083
Total cost$0.2625

Priced from 2026-03-16 · Input $/M $5.00 · Output $/M $25.00

Result 134 of 177

Test T0687 at 2026-03-16

manuscript transcription handwritten 15 de prose

genai gemini-3.1-pro-preview · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score98.2%
CER ↓0.9%

{"folios":[{"folio":"3","text":"Vnd ein pferit die mir vnd
minen knechten vber hulfend
den do was nienan kein weg
denne den wir machtend
vnd vielend die knecht dik
vnd vil in vntz an den ars
vnd die pferit vntz an die
settel vnd was ze mol ein grosser
nebel dz wir kum gesachend
vnd also mit grosser arbeit kome
wir ze mittem tag zu sant
kristoffel vff den berg do
do sach ich die bucher do gar
vil herren wopen jn stont
die ir stur do hin geben hand
do stund mines vatters seligen
wopen och in dem einen","addition1":"","addition2":"","addition3":""}]}

Scoring
Speed

Model time, all inputs543.9 s
Mean per input49.45 s
Slowest input458.12 s
Inputs timed11

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

No cost recorded for this run

Priced from 2026-03-16 · Input $/M $2.00 · Output $/M $12.00

Result 135 of 177

Test T0658 at 2026-03-16

manuscript transcription handwritten 15 de prose

openai gpt-5.4-2026-03-05 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score92.7%
CER ↓7.4%

{"folios":[{"folio":"3","text":"Vnd ein pferit die mir mit
minen knechten vber hulffend
bay do was meinen kẽy weg
Sünni. den wir machtenß
vnd irslend die knecht druck
vnd vil in vntz an den aͤrß
vnd die pferit vntz an die
petel vnd was ze mol ein grosser
nebel dz wir kaum geꝑaͤchend
vnd alß mit groffer arbeit kam
wir ze mitteng tag zu sant
kristoffel vff den berg do
do hub ich die bücher do gar
vil hering wappen zu stont
Die ir für delfin geling hand
Do fund mines vatters flügen
wappen och in dem einen","addition1":"","addition2":"","addition3":""}]}

Scoring

Fuzzy score76.5%
CER ↓26.4%

Speed

Model time, all inputs95.5 s
Mean per input7.96 s
Slowest input13.05 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 tokens40.2K
Output tokens2.6K
Total tokens42.9K
Input cost$0.1006
Output cost$0.0398
Total cost$0.1403

Priced from 2026-03-16 · Input $/M $2.50 · Output $/M $15.00

Result 136 of 177

Test T0646 at 2026-03-16

manuscript transcription handwritten 15 de prose

anthropic claude-sonnet-4-6 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score97.5%
CER ↓1.7%

{"folios":[{"folio":"3","text":"Vnd ein pferit die mir vnd
minen knechten vber hulfend
den do was menen kein weg
denne den wir machtend
vnd viehend die knecht dick
vnd vil in vntz an den arß
vnd die pferit vntz an die
settel vnd was ze mol ein grosse
nebel dz wir kum gesahend
vnd alß mit grosser arbeit kame
wir ze mittem tag zu sant
kristoffel vff den berg do
do sach ich die bücher do gar
vil herren wopen in stone
die ir stür do hin geben hand
do stund mines vatters seligen
wopen och in dem einen","addition1":null,"addition2":null,"addition3":null}]}

Scoring

Fuzzy score75.9%
CER ↓23.9%

Speed

Model time, all inputs136.8 s
Mean per input11.40 s
Slowest input17.09 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 tokens35.9K
Output tokens3.3K
Total tokens39.3K
Input cost$0.1077
Output cost$0.0502
Total cost$0.1579

Priced from 2026-03-16 · Input $/M $3.00 · Output $/M $15.00

Result 137 of 177

Test T0538 at 2026-02-18

manuscript transcription handwritten 15 de prose

mistral mistral-large-2512 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score0.0%
CER ↓100.0%

{"folios":[{"folio":"","text":"","addition1":"","addition2":"","addition3":""}]}

Scoring

Fuzzy score0.0%
CER ↓100.0%

Speed

Model time, all inputs16.0 s
Mean per input1.34 s
Slowest input3.47 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 tokens9K
Output tokens552
Total tokens9.6K
Input cost$0.0045
Output cost$0.0008
Total cost$0.0053

Priced from 2026-01-23 · Input $/M $0.50 · Output $/M $1.50

Result 138 of 177

Test T0276 at 2026-02-17

manuscript transcription handwritten 15 de prose

openai gpt-4o-mini-2024-07-18 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score89.2%
CER ↓13.2%

{"folios":[{"folio":"3","text":"Und ein pferit die mir mit
 meinen knahten vnder Rüffend
 das so was menner künig weg
 same das mir machtend
 und vieland die knecht dut
 und vil in vntz von eyn arb
 und die pffent vng sy die
 petal und was ze mol ein groser
 nebel ob wir kum gepalt
 und als mit grosser arbeit komme
 wir ze mittem tag zu rnt
 einstolff vff day derg do
 so fägt nog die büger do gar
 mit herrn, wappen yn bon
 die in fier do fin geden grand
 so fünd minnes vatters plagen
 wappen oft in sem eyn","addition1":"","addition2":"","addition3":""}]}

Scoring

Fuzzy score56.2%
CER ↓48.3%

Speed

Model time, all inputs107.7 s
Mean per input8.97 s
Slowest input12.11 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 tokens450K
Output tokens2.4K
Total tokens452.4K
Input cost$0.0675
Output cost$0.0015
Total cost$0.069

Priced from 2026-01-23 · Input $/M $0.15 · Output $/M $0.60

Result 139 of 177

Test T0278 at 2026-02-17

manuscript transcription handwritten 15 de prose

openai gpt-4.1-nano-2025-04-14 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score87.2%
CER ↓15.8%

{"folios":[{"folio":"3","text":"Vnd ein pesserit die mir vnd
 kinnet encoghen vnder gultend
 den so nur mîner zeyt vnd
 dame den mir machent
 vnd vichent die encoge gut
 vnd vil in vnger den arb
 vnd die pesserit vnd my die
 stedel vnd war ze mol in grosse
 nebet ez wir kum geschehen
 vnd als mit grosser arbei kam
 vnm ze mittem tag zu bote
 einstolff uff den dant so
 do so alt ez die brüder do jar
 vil herring vnspey zu sone
 die in sÿr delÿn gebÿg hand
 do pount mîner vater vÿn
 vnspey och in dem eÿn

","addition1":"","addition2":"","addition3":""}]}

Scoring

Fuzzy score53.8%
CER ↓51.6%

Speed

Model time, all inputs734.7 s
Mean per input61.23 s
Slowest input382.18 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 tokens51.1K
Output tokens2.2K
Total tokens53.3K
Input cost$0.0051
Output cost$0.0009
Total cost$0.006

Priced from 2026-01-23 · Input $/M $0.10 · Output $/M $0.40

Result 140 of 177

Test T0409 at 2026-01-26

manuscript transcription handwritten 15 de prose

openai gpt-5.1-2025-11-13 · temp 0.0 · dataclass Document

Response vs ground truth

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

Fuzzy score89.5%
CER ↓11.8%

{"folios":[{"folio":"3","text":"vnd ein pferit die mir vnd
minen knechten vber hulffend
das So was miners tänt weg
same das wir machtend
und vilclach die knecht dut
und vil in vntz an den ardt
und die pferit vntz an die
statl und was ze mal ein groszer
nebel das wir kun geſehend
und als mit groszer arbait kamen
wir ze mitten tag zů sant
frantzstol vff den berg So
So sach ich die buͤrger So gar
vil herren wapẝen zu pferd
vnd in aller scheyn gestenck gantz
vñ sinnd miners vatters flagen
wapñ auch in dem schein","addition1":"","addition2":"","addition3":""}]}

Scoring

Fuzzy score65.5%
CER ↓37.1%

Speed

Model time, all inputs102.0 s
Mean per input8.50 s
Slowest input12.26 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 tokens18.8K
Output tokens2.6K
Total tokens21.5K
Input cost$0.0235
Output cost$0.0263
Total cost$0.0498

Priced from 2026-01-23 · Input $/M $1.25 · Output $/M $10.00