Benchmark Almanac

PeerQA

Mistral-7B-Instruct-v0.2 0.24 leads since February 2025
Not in model cards

Models answer peer reviewers' questions about a scientific article using evidence from the article.

The original authors supplied reference answers to the reviewers' questions.

Made by
UKP Lab
Released
February 2025
Kind
Research science
Level
Expert
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Gemma 3 27B #165 on the capability index
Last model added
May 2025
Scores refreshed
November 20th, 2026
14
Entries in its paper, 10 models we track

Mistral-7B-Instruct-v0.2 leads with 0.24 among models released by May 2025 Mistral-7B-Instruct-v0.2 leads with 0.24 among models released by May 2025

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The headline is answer-generation ROUGE-L against the authors’ answer evidence, the first measure in Table 12. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in May 2025, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Mistral-7B-Instruct-v0.2 Gold evidence 0.24
  2. 2 GPT-4o Aug 2024, Gold evidence 0.20
  3. 3 Llama-3 IT-8B-32k Full text 0.18
  4. 4 Llama 3.3 70B Instruct Full text 0.18
  5. 5 Llama-3 IT-8B-8k Top 20 paragraphs 0.18
  6. 6 deepseek-r1-0528-qwen3-8b Gold evidence 0.17
  7. 7 DeepSeek-R1-Llama-8B-128k Gold evidence 0.17
  8. 8 DeepSeek-R1-Distill-Qwen-32B Gold evidence 0.17
  9. 9 DeepSeek-R1-Distill-Qwen-14B Gold evidence 0.16
  10. 10 DeepSeek-R1-Distill-Qwen-7B Top 20 paragraphs 0.16
  11. 11 Gemma 3 27B Full text 0.16
  12. 12 GPT-3.5 Turbo Jun 2023, Gold evidence 0.15
  13. 13 DeepSeek-R1-Distill-Llama-70B Gold evidence 0.15
  14. 14 Command-R v01-34B-128k Gold evidence 0.13

Source: The PeerQA paper

#EntryScoreSettingDateSource
1 Mistral-7B-Instruct-v0.2 0.24 Gold evidence Dec 2023 Official leaderboard
2 GPT-4o Aug 2024 0.20 Gold evidence Aug 2024 Official leaderboard
3 Llama-3 IT-8B-32k 0.18 Full text Official leaderboard
4 Llama 3.3 70B Instruct 0.18 Full text Dec 2024 Official leaderboard
5 Llama-3 IT-8B-8k 0.18 Top 20 paragraphs Official leaderboard
6 deepseek-r1-0528-qwen3-8b 0.17 Gold evidence May 2025 Official leaderboard
7 DeepSeek-R1-Llama-8B-128k 0.17 Gold evidence Official leaderboard
8 DeepSeek-R1-Distill-Qwen-32B 0.17 Gold evidence Jan 2025 Official leaderboard
9 DeepSeek-R1-Distill-Qwen-14B 0.16 Gold evidence Jan 2025 Official leaderboard
10 DeepSeek-R1-Distill-Qwen-7B 0.16 Top 20 paragraphs Jan 2025 Official leaderboard
11 Gemma 3 27B 0.16 Full text Mar 2025 Official leaderboard
12 GPT-3.5 Turbo Jun 2023 0.15 Gold evidence Jun 2023 Official leaderboard
13 DeepSeek-R1-Distill-Llama-70B 0.15 Gold evidence Jan 2025 Official leaderboard
14 Command-R v01-34B-128k 0.13 Gold evidence Official leaderboard
#ModelScoreSettingDateSource
1 Mistral-7B-Instruct-v0.2 0.24 Gold evidence Dec 2023 Official leaderboard
2 GPT-4o Aug 2024 0.20 Gold evidence Aug 2024 Official leaderboard
3 Llama 3.3 70B Instruct 0.18 Full text Dec 2024 Official leaderboard
4 deepseek-r1-0528-qwen3-8b 0.17 Gold evidence May 2025 Official leaderboard
5 DeepSeek-R1-Distill-Qwen-32B 0.17 Gold evidence Jan 2025 Official leaderboard
6 DeepSeek-R1-Distill-Qwen-14B 0.16 Gold evidence Jan 2025 Official leaderboard
7 DeepSeek-R1-Distill-Qwen-7B 0.16 Top 20 paragraphs Jan 2025 Official leaderboard
8 Gemma 3 27B 0.16 Full text Mar 2025 Official leaderboard
9 GPT-3.5 Turbo Jun 2023 0.15 Gold evidence Jun 2023 Official leaderboard
10 DeepSeek-R1-Distill-Llama-70B 0.15 Gold evidence Jan 2025 Official leaderboard

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Mistral-7B-Instruct-v0.2
Full text Top 50 para… Gold eviden…
0.24
GPT-4o
Top 10 para… Gold eviden…
0.20
Llama 3.3 70B Instruct
Gold eviden… Full text
0.18
deepseek-r1-0528-qwen3-8b
Top 10 para… Gold eviden…
0.17
DeepSeek-R1-Distill-Qwen-32B
Top 10 para… Gold eviden…
0.17
DeepSeek-R1-Distill-Qwen-14B
Top 10 para… Gold eviden…
0.16
DeepSeek-R1-Distill-Qwen-7B
Full text Top 20 para…
0.16
Gemma 3 27B
Top 10 para… Full text
0.16
GPT-3.5 Turbo
Full text Top 100 par… Gold eviden…
0.15
DeepSeek-R1-Distill-Llama-70B
Top 10 para… Full text
0.15

Source: The PeerQA paper

ModelEffortHarnessScoreCost per taskOutput tokens
Mistral-7B-Instruct-v0.2 Gold evidence 0.24
Mistral-7B-Instruct-v0.2 Top 20 paragraphs 0.20
Mistral-7B-Instruct-v0.2 Top 50 paragraphs 0.20
Mistral-7B-Instruct-v0.2 Top 100 paragraphs 0.20
GPT-4o Aug 2024 Gold evidence 0.20
Mistral-7B-Instruct-v0.2 Top 10 paragraphs 0.20
Mistral-7B-Instruct-v0.2 Full text 0.19
GPT-4o Aug 2024 Full text 0.18
Llama 3.3 70B Instruct Full text 0.18
GPT-4o Aug 2024 Top 20 paragraphs 0.18
GPT-4o Aug 2024 Top 100 paragraphs 0.18
GPT-4o Aug 2024 Top 50 paragraphs 0.18
GPT-4o Aug 2024 Top 10 paragraphs 0.18
deepseek-r1-0528-qwen3-8b Gold evidence 0.17
DeepSeek-R1-Distill-Qwen-32B Gold evidence 0.17
DeepSeek-R1-Distill-Qwen-14B Gold evidence 0.16
DeepSeek-R1-Distill-Qwen-32B Top 50 paragraphs 0.16
Llama 3.3 70B Instruct Top 10 paragraphs 0.16
DeepSeek-R1-Distill-Qwen-32B Top 100 paragraphs 0.16
DeepSeek-R1-Distill-Qwen-32B Full text 0.16
DeepSeek-R1-Distill-Qwen-7B Top 20 paragraphs 0.16
Gemma 3 27B Full text 0.16
deepseek-r1-0528-qwen3-8b Top 100 paragraphs 0.16
DeepSeek-R1-Distill-Qwen-14B Top 100 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-32B Top 20 paragraphs 0.15
deepseek-r1-0528-qwen3-8b Top 20 paragraphs 0.15
GPT-3.5 Turbo Jun 2023 Gold evidence 0.15
DeepSeek-R1-Distill-Llama-70B Gold evidence 0.15
DeepSeek-R1-Distill-Qwen-14B Full text 0.15
DeepSeek-R1-Distill-Llama-70B Full text 0.15
DeepSeek-R1-Distill-Qwen-14B Top 50 paragraphs 0.15
deepseek-r1-0528-qwen3-8b Top 50 paragraphs 0.15
Llama 3.3 70B Instruct Gold evidence 0.15
deepseek-r1-0528-qwen3-8b Top Full Text paragraphs 0.15
DeepSeek-R1-Distill-Qwen-32B Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Top 50 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-14B Top 20 paragraphs 0.15
deepseek-r1-0528-qwen3-8b Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-14B Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Top 100 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Gold evidence 0.14
DeepSeek-R1-Distill-Qwen-7B Full text 0.14
DeepSeek-R1-Distill-Llama-70B Top 10 paragraphs 0.14
Gemma 3 27B Gold evidence 0.14
GPT-3.5 Turbo Jun 2023 Top 20 paragraphs 0.14
GPT-3.5 Turbo Jun 2023 Top 50 paragraphs 0.14
GPT-3.5 Turbo Jun 2023 Top 10 paragraphs 0.13
Gemma 3 27B Top 10 paragraphs 0.13
GPT-3.5 Turbo Jun 2023 Top 100 paragraphs 0.13
GPT-3.5 Turbo Jun 2023 Full text 0.12

Scores at each effort level

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Mistral-7B-Instruct-v0.2
Full text Top 50 para… Gold eviden…
0.24
GPT-4o
Top 10 para… Gold eviden…
0.20
Llama 3.3 70B Instruct
Gold eviden… Full text
0.18
deepseek-r1-0528-qwen3-8b
Top 10 para… Gold eviden…
0.17
DeepSeek-R1-Distill-Qwen-32B
Top 10 para… Gold eviden…
0.17
DeepSeek-R1-Distill-Qwen-14B
Top 10 para… Gold eviden…
0.16
DeepSeek-R1-Distill-Qwen-7B
Full text Top 20 para…
0.16
Gemma 3 27B
Top 10 para… Full text
0.16
GPT-3.5 Turbo
Full text Top 100 par… Gold eviden…
0.15
DeepSeek-R1-Distill-Llama-70B
Top 10 para… Full text
0.15

Source: The PeerQA paper

ModelEffortHarnessScoreCost per taskOutput tokens
Mistral-7B-Instruct-v0.2 Gold evidence 0.24
Mistral-7B-Instruct-v0.2 Top 20 paragraphs 0.20
Mistral-7B-Instruct-v0.2 Top 50 paragraphs 0.20
Mistral-7B-Instruct-v0.2 Top 100 paragraphs 0.20
GPT-4o Aug 2024 Gold evidence 0.20
Mistral-7B-Instruct-v0.2 Top 10 paragraphs 0.20
Mistral-7B-Instruct-v0.2 Full text 0.19
GPT-4o Aug 2024 Full text 0.18
Llama 3.3 70B Instruct Full text 0.18
GPT-4o Aug 2024 Top 20 paragraphs 0.18
GPT-4o Aug 2024 Top 100 paragraphs 0.18
GPT-4o Aug 2024 Top 50 paragraphs 0.18
GPT-4o Aug 2024 Top 10 paragraphs 0.18
deepseek-r1-0528-qwen3-8b Gold evidence 0.17
DeepSeek-R1-Distill-Qwen-32B Gold evidence 0.17
DeepSeek-R1-Distill-Qwen-14B Gold evidence 0.16
DeepSeek-R1-Distill-Qwen-32B Top 50 paragraphs 0.16
Llama 3.3 70B Instruct Top 10 paragraphs 0.16
DeepSeek-R1-Distill-Qwen-32B Top 100 paragraphs 0.16
DeepSeek-R1-Distill-Qwen-32B Full text 0.16
DeepSeek-R1-Distill-Qwen-7B Top 20 paragraphs 0.16
Gemma 3 27B Full text 0.16
deepseek-r1-0528-qwen3-8b Top 100 paragraphs 0.16
DeepSeek-R1-Distill-Qwen-14B Top 100 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-32B Top 20 paragraphs 0.15
deepseek-r1-0528-qwen3-8b Top 20 paragraphs 0.15
GPT-3.5 Turbo Jun 2023 Gold evidence 0.15
DeepSeek-R1-Distill-Llama-70B Gold evidence 0.15
DeepSeek-R1-Distill-Qwen-14B Full text 0.15
DeepSeek-R1-Distill-Llama-70B Full text 0.15
DeepSeek-R1-Distill-Qwen-14B Top 50 paragraphs 0.15
deepseek-r1-0528-qwen3-8b Top 50 paragraphs 0.15
Llama 3.3 70B Instruct Gold evidence 0.15
deepseek-r1-0528-qwen3-8b Top Full Text paragraphs 0.15
DeepSeek-R1-Distill-Qwen-32B Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Top 50 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-14B Top 20 paragraphs 0.15
deepseek-r1-0528-qwen3-8b Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-14B Top 10 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Top 100 paragraphs 0.15
DeepSeek-R1-Distill-Qwen-7B Gold evidence 0.14
DeepSeek-R1-Distill-Qwen-7B Full text 0.14
DeepSeek-R1-Distill-Llama-70B Top 10 paragraphs 0.14
Gemma 3 27B Gold evidence 0.14
GPT-3.5 Turbo Jun 2023 Top 20 paragraphs 0.14
GPT-3.5 Turbo Jun 2023 Top 50 paragraphs 0.14
GPT-3.5 Turbo Jun 2023 Top 10 paragraphs 0.13
Gemma 3 27B Top 10 paragraphs 0.13
GPT-3.5 Turbo Jun 2023 Top 100 paragraphs 0.13
GPT-3.5 Turbo Jun 2023 Full text 0.12

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Mistral AI
Benchmark paper Not tested since May 2025

Source: The PeerQA paper

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper
0.10 0.15 0.20 0.25 2023 2024 2025 0.10 0.15 0.20 0.25 2023 2024 2025

Source: Official PeerQA leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Science and reasoning

All 203 in Science and reasoning
  • GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics. Gemini 4 Carbon 97%
  • Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
  • BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
  • FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics. Intern-S2-397B 87%
  • HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels. Gemini 3 Pro 78.3%
  • LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 18 months ago, in May 2025, for deepseek-r1-0528-qwen3-8b. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the benchmark’s paper (14 entries).
  • Pass Models scored 10 models we track, plus 4 other entries on its leaderboard.
  • Warning Newest score The newest came 18 months ago, in May 2025, for deepseek-r1-0528-qwen3-8b.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Gemma 3 27B (#165).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 51 rows for models we track and 4 other entries at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check