AstroVisBench
Models write code to process astronomical observations and plot the results. The benchmark compares computed values with reference outputs and uses a model to grade plots.
The authors publish seven metrics for eight models. The 864 tasks come from 110 astronomy notebooks.
- Made by
- University of Texas at Austin and collaborators
- Released
- May 2025
- Kind
- Lab and data work
- Level
- University
- Scored
- Graded by a model
- Answers
- Code
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Single turn
- Environment
- Code sandbox
- Items from
- Real work and use
- Refresh
- One fixed set
- Run by
- Anyone
- Strongest model tested
- Claude Opus 4 #109 on the capability index
- Last model added
- May 2025
- Scores refreshed
- November 20th, 2026
Gemini 2.5 Pro (snapshot unspecified) tops the leaderboard with 69.2% Claude Opus 4 leads with 49.8% among models released by May 2025
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Claude Opus 4, at 49.8%. 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:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 Gemini 2.5 Pro (snapshot unspecified) 69.2%
- 2 Claude Opus 4 49.8%
- 3 Claude 3.7 Sonnet 49.1%
- 4 o3-mini 48.6%
- 5 GPT-4o Mar 2025 46.3%
- 6
Llama 4 Maverick 44.7%
- 7
Qwen2.5-72B 43.1%
- 8
QwQ-32B 35.9%
Source: Official AstroVisBench leaderboard
- 1 Claude Opus 4 49.8%
- 2 Claude 3.7 Sonnet 49.1%
- 3 o3-mini 48.6%
- 4 GPT-4o Mar 2025 46.3%
- 5
Llama 4 Maverick 44.7%
- 6
Qwen2.5-72B 43.1%
- 7
QwQ-32B 35.9%
Source: Official AstroVisBench leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Gemini 2.5 Pro (snapshot unspecified) | 69.2% | Official leaderboard | ||
| 2 | Claude Opus 4 | 49.8% | May 2025 | Official leaderboard | |
| 3 | Claude 3.7 Sonnet | 49.1% | Feb 2025 | Official leaderboard | |
| 4 | o3-mini | 48.6% | Jan 2025 | Official leaderboard | |
| 5 | GPT-4o Mar 2025 | 46.3% | Mar 2025 | Official leaderboard | |
| 6 | | 44.7% | Apr 2025 | Official leaderboard | |
| 7 | | 43.1% | Sep 2024 | Official leaderboard | |
| 8 | | 35.9% | Mar 2025 | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4 | 49.8% | May 2025 | Official leaderboard | |
| 2 | Claude 3.7 Sonnet | 49.1% | Feb 2025 | Official leaderboard | |
| 3 | o3-mini | 48.6% | Jan 2025 | Official leaderboard | |
| 4 | GPT-4o Mar 2025 | 46.3% | Mar 2025 | Official leaderboard | |
| 5 | | 44.7% | Apr 2025 | Official leaderboard | |
| 6 | | 43.1% | Sep 2024 | Official leaderboard | |
| 7 | | 35.9% | Mar 2025 | Official leaderboard |
Rankings by part
We read 7 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as NoErr(P) %, VIscore and NoErr(V) %. o3-mini leads 2 of them, and 4 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
NoErr(P) %
7 models we track have a score here.
- 1 Claude Opus 4 49.8%
- 2 Claude 3.7 Sonnet 49.1%
- 3 o3-mini 48.6%
- 4 GPT-4o Mar 2025 46.3%
- 5
Llama 4 Maverick 44.7%
- 6
Qwen2.5-72B 43.1%
- 7
QwQ-32B 35.9%
VIscore
7 models we track have a score here.
- 1 o3-mini ↑2 0.69
- 2 Claude Opus 4 ↓1 0.64
- 3 Claude 3.7 Sonnet ↓1 0.63
- 4
Llama 4 Maverick ↑1 0.55
- 5
Qwen2.5-72B ↑1 0.53
- 6 GPT-4o Mar 2025 ↓2 0.48
- 7
QwQ-32B 0.47
NoErr(V) %
7 models we track have a score here.
- 1
Llama 4 Maverick ↑4 71.3%
- 2 o3-mini ↑1 69.7%
- 3 GPT-4o Mar 2025 ↑1 67.4%
- 4 Claude Opus 4 ↓3 66.9%
- 5 Claude 3.7 Sonnet ↓3 65.3%
- 6
Qwen2.5-72B 64.6%
- 7
QwQ-32B 39.1%
CorrectV %
7 models we track have a score here.
- 1
Qwen2.5-72B ↑5 10.4%
- 2 o3-mini ↑1 10.4%
- 3
Llama 4 Maverick ↑2 9.70%
- 4 Claude 3.7 Sonnet ↓2 9.50%
- 5 Claude Opus 4 ↓4 9.30%
- 6 GPT-4o Mar 2025 ↓2 8.60%
- 7
QwQ-32B 8.20%
VisFail %
7 models we track have a score here.
- 1 o3-mini ↑2 2.80%
- 2
Qwen2.5-72B ↑4 3.70%
- 3
QwQ-32B ↑4 3.90%
- 4 GPT-4o Mar 2025 6.00%
- 5
Llama 4 Maverick 9.00%
- 6 Claude Opus 4 ↓5 12.7%
- 7 Claude 3.7 Sonnet ↓5 21.3%
MiE %
7 models we track have a score here.
- 1
QwQ-32B ↑6 12.6%
- 2 Claude 3.7 Sonnet 14.6%
- 3
Qwen2.5-72B ↑3 20.1%
- 4
Llama 4 Maverick ↑1 21.5%
- 5 GPT-4o Mar 2025 ↓1 23.4%
- 6 o3-mini ↓3 26.4%
- 7 Claude Opus 4 ↓6 28.9%
MaE %
7 models we track have a score here.
- 1
QwQ-32B ↑6 14.9%
- 2 Claude Opus 4 ↓1 16%
- 3 Claude 3.7 Sonnet ↓1 19.2%
- 4 GPT-4o Mar 2025 28.5%
- 5 o3-mini ↓2 29.6%
- 6
Qwen2.5-72B 29.9%
- 7
Llama 4 Maverick ↓2 30.6%
2 models have held number one on the official leaderboard since its release in May 2025
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official AstroVisBench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official AstroVisBench 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
- 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
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 18 months ago, in May 2025, for Claude Opus 4. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the official leaderboard (8 entries).
- Pass Models scored 7 models we track, plus 1 other entry on its leaderboard.
- Warning Newest score The newest came 18 months ago, in May 2025, for Claude Opus 4.
- Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 4 (#109).
- Pass Headroom The best is 49.8%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on astrovisbench.github.io: 7 rows for models we track and 1 other entry 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.