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.

Search Results
Show compact results Refine Search New Search
Download JSON Download CSV (Downloads are limited to 250 results.)

Your search for Benchmark 'business_letters__true' with Search Hidden 'False' returned 613 results, showing page 3 of 62.
Result 21 of 613

Test T1856 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

x-ai grok-4.7 · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp1
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":null,"sender_persons":["Basler Rheinschiffahrt-Aktiengesellschaft"],"receiver_persons":["Christ"]}}

Scoring

F1 micro66.0%
F1 macro64.0%

Speed

Model time, all inputs2915.6 s
Mean per input51.15 s
Slowest input240.88 s
Inputs timed57

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 tokens340.4K
Output tokens3K
Total tokens343.4K
Input cost$0.6807
Output cost$0.0182
Total cost$1.91
Reasoning cost usd$1.21
Total reasoning tokens202.3K

Priced from 2026-09-27 · Input $/M $2.00 · Output $/M $6.00

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 22 of 613

Test T1858 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

x-ai grok-4.7 · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp1
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":null,"sender_persons":["Basler Rheinschiffahrt-Aktiengesellschaft"],"receiver_persons":["Christ"]}}

Scoring

F1 micro58.0%
F1 macro62.0%

Speed

Model time, all inputs1654.2 s
Mean per input51.69 s
Slowest input156.79 s
Inputs timed32

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 tokens210.4K
Output tokens2K
Total tokens212.4K
Input cost$0.4207
Output cost$0.0121
Total cost$1.26
Reasoning cost usd$0.826
Total reasoning tokens137.7K

Priced from 2026-09-27 · Input $/M $2.00 · Output $/M $6.00

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 23 of 613

Test T1841 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

openai gpt-6-luna · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp1
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":["Zusammenarbeit im Personenschifffahrtsgeschäft"],"sender_persons":["E. Müller"],"receiver_persons":["Christ"]}}

Scoring

F1 micro62.0%
F1 macro59.0%

Speed

Model time, all inputs273.0 s
Mean per input4.79 s
Slowest input9.36 s
Inputs timed57

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 tokens374K
Output tokens18.4K
Total tokens392.3K
Input cost$0.0374
Output cost$0.0092
Total cost$0.0466

Priced from 2026-09-27 · Input $/M $0.10 · Output $/M $0.50

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 24 of 613

Test T1813 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

anthropic claude-opus-5-5 · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn1
Sender persons fp1
Sender persons tp1

```json
{
    "letter_title": ["null"],
    "send_date": ["1926-02-16"],
    "sender_persons": ["Groschupf", "F. Ritter"],
    "receiver_persons": ["Christ"]
}
```

Scoring

F1 micro67.0%
F1 macro71.0%

Speed

Model time, all inputs207.1 s
Mean per input6.47 s
Slowest input12.22 s
Inputs timed32

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 tokens257.8K
Output tokens10.9K
Total tokens268.7K
Input cost$1.03
Output cost$0.2179
Total cost$1.25

Priced from 2026-09-27 · Input $/M $4.00 · Output $/M $20.00

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 25 of 613

Test T1812 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

anthropic claude-opus-5-5 · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn0
Sender persons fp0
Sender persons tp0

{
    "letter_title": ["null"],
    "send_date": ["1926-03-29"],
    "sender_persons": ["null"],
    "receiver_persons": ["Dr. W. Krasting"]
}

Scoring

F1 micro87.0%
F1 macro76.0%

Speed

Model time, all inputs130.6 s
Mean per input5.22 s
Slowest input11.72 s
Inputs timed25

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 tokens146.7K
Output tokens6.5K
Total tokens153.1K
Input cost$0.5867
Output cost$0.1291
Total cost$0.7158

Priced from 2026-09-27 · Input $/M $4.00 · Output $/M $20.00

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 26 of 613

Test T1826 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

openai gpt-6-sol · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp1
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":null,"sender_persons":["P. Müller"],"receiver_persons":["Herr Christ"]}}

Scoring

F1 micro70.0%
F1 macro66.0%

Speed

Model time, all inputs310.6 s
Mean per input5.45 s
Slowest input10.80 s
Inputs timed57

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 tokens374K
Output tokens13K
Total tokens387K
Input cost$0.7479
Output cost$0.1301
Total cost$0.878

Priced from 2026-09-27 · Input $/M $2.00 · Output $/M $10.00

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 27 of 613

Test T1843 at 2026-09-27

letter information-extraction typed, handwritten 20 de prose person, location

openai gpt-6-luna · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp0
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":["null"],"sender_persons":["null"],"receiver_persons":["Christ"]}}

Scoring

F1 micro61.0%
F1 macro63.0%

Speed

Model time, all inputs165.3 s
Mean per input5.17 s
Slowest input8.19 s
Inputs timed32

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 tokens238.1K
Output tokens11.2K
Total tokens249.3K
Input cost$0.0238
Output cost$0.0056
Total cost$0.0294

Priced from 2026-09-27 · Input $/M $0.10 · Output $/M $0.50

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 28 of 613

Test T1707 at 2026-09-18

letter information-extraction typed, handwritten 20 de prose person, location

openrouter z-ai/glm-5.3-flash · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn0
Sender persons fp0
Sender persons tp0

{"metadata":{"send_date":["1926-03-29"],"letter_title":["null"],"sender_persons":["null"],"receiver_persons":["Dr. W. Krasting"]}}

Scoring

F1 micro62.0%
F1 macro53.0%

Speed

Model time, all inputs124.4 s
Mean per input4.98 s
Slowest input24.21 s
Inputs timed25

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 tokens150.2K
Output tokens6.7K
Total tokens156.9K
Input cost$0.0113
Output cost$0.0017
Total cost$0.0129

Priced from 2026-09-08 · Input $/M $0.075 · Output $/M $0.25

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 29 of 613

Test T1706 at 2026-09-18

letter information-extraction typed, handwritten 20 de prose person, location

openrouter z-ai/glm-5.3-flash · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp2
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":["null"],"sender_persons":["Gyssling","Dr. Rilliet"],"receiver_persons":["Herr Christ"]}}

Scoring

F1 micro58.0%
F1 macro57.0%

Speed

Model time, all inputs2433.1 s
Mean per input43.45 s
Slowest input302.99 s
Inputs timed56

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 tokens381.3K
Output tokens112.1K
Total tokens493.4K
Input cost$0.0307
Output cost$0.0302
Total cost$0.0609

Priced from 2026-09-08 · Input $/M $0.075 · Output $/M $0.25

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827
Result 30 of 613

Test T1708 at 2026-09-18

letter information-extraction typed, handwritten 20 de prose person, location

openrouter z-ai/glm-5.3-flash · temp 0.0 · dataclass Document

Response vs ground truth

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

Receiver persons fn0
Receiver persons fp0
Receiver persons tp1
Send date fn0
Send date fp0
Send date tp1
Sender persons fn2
Sender persons fp1
Sender persons tp0

{"metadata":{"send_date":["1926-02-16"],"letter_title":[],"sender_persons":["Basler Rheinschiffahrt-Aktiengesellschaft"],"receiver_persons":["Herr Christ"]}}

Scoring

F1 micro55.0%
F1 macro60.0%

Speed

Model time, all inputs507.7 s
Mean per input15.87 s
Slowest input388.55 s
Inputs timed32

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 tokens263.7K
Output tokens3.8K
Total tokens267.5K
Input cost$0.0203
Output cost$0.0011
Total cost$0.0215

Priced from 2026-09-08 · Input $/M $0.075 · Output $/M $0.25

Cite: Hindermann, Marti, Alberto, et al., (2026). RISE-UNIBAS/humanities_data_benchmark, 10.5281/zenodo.21975827