Benchmark Almanac

MatCha

Claude-3.5-Sonnet (snapshot unspecified) 68.6% tops its leaderboard Llama 4 Maverick 68.1% leads the models we track
Not in model cards

Models answer multiple-choice questions about material structure, shape and properties using images from materials characterization instruments.

The authors provide 1,500 questions across 21 tasks. The ATOMIC team independently compares models on the transmission electron microscopy subset.

Released
September 2025
Kind
Lab and data work
Level
Expert
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Images → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
Occupations
Life, physical and social science
Strongest model tested
Llama 4 Maverick #153 on the capability index
Last model added
April 2025
Scores refreshed
November 20th, 2026
15
Entries in its paper, 9 models we track

Claude-3.5-Sonnet (snapshot unspecified) tops the leaderboard with 68.6% Llama 4 Maverick leads with 68.1% among models released by April 2025

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. 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. The best model we track is Llama 4 Maverick, at 68.1%. This view shows only models we track, each at its best.

The newest model with a score here came out in April 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 Claude-3.5-Sonnet (snapshot unspecified) 8-shot 68.6%
  2. 2 Llama 4 Maverick 16-shot 68.1%
  3. 3 GPT-4o (snapshot unspecified) 16-shot 68%
  4. 4 Gemini-1.5-Pro (snapshot unspecified) 4-shot 57.1%
  5. 5 Gemini-1.5-Flash (snapshot unspecified) 16-shot 52.3%
  6. 6 InternVL3-38B 2-shot 52.2%
  7. 7 GPT-4o-mini (snapshot unspecified) 16-shot 50.1%
  8. 8 Qwen2.5-VL-32B Chain of thought 50.1%
  9. 9 InternVL3-8B 8-shot 44.6%
  10. 10 Qwen2.5-VL-7B Chain of thought 42.7%
  11. 11 Janus-Pro-7B Chain of thought 40.3%
  12. 12 Gemma 3 4B 8-shot 35.1%
  13. 13 Llama-3.2-11B-Vision Chain of thought 34.8%
  14. 14 LLaVA-v1.5-13B 2-shot 29.7%
  15. 15 llava-v1.5-7b 2-shot 24.1%

Source: The MatCha paper

#EntryScoreSettingDateSource
1 Claude-3.5-Sonnet (snapshot unspecified) 68.6% 8-shot Official leaderboard
2 Llama 4 Maverick 68.1% 16-shot Apr 2025 Official leaderboard
3 GPT-4o (snapshot unspecified) 68% 16-shot Official leaderboard
4 Gemini-1.5-Pro (snapshot unspecified) 57.1% 4-shot Official leaderboard
5 Gemini-1.5-Flash (snapshot unspecified) 52.3% 16-shot Official leaderboard
6 InternVL3-38B 52.2% 2-shot Apr 2025 Official leaderboard
7 GPT-4o-mini (snapshot unspecified) 50.1% 16-shot Official leaderboard
8 Qwen2.5-VL-32B 50.1% Chain of thought Mar 2025 Official leaderboard
9 InternVL3-8B 44.6% 8-shot Apr 2025 Official leaderboard
10 Qwen2.5-VL-7B 42.7% Chain of thought Jan 2025 Official leaderboard
11 Janus-Pro-7B 40.3% Chain of thought Jan 2025 Official leaderboard
12 Gemma 3 4B 35.1% 8-shot Mar 2025 Official leaderboard
13 Llama-3.2-11B-Vision 34.8% Chain of thought Official leaderboard
14 LLaVA-v1.5-13B 29.7% 2-shot Oct 2023 Official leaderboard
15 llava-v1.5-7b 24.1% 2-shot Oct 2023 Official leaderboard
#ModelScoreSettingDateSource
1 Llama 4 Maverick 68.1% 16-shot Apr 2025 Official leaderboard
2 InternVL3-38B 52.2% 2-shot Apr 2025 Official leaderboard
3 Qwen2.5-VL-32B 50.1% Chain of thought Mar 2025 Official leaderboard
4 InternVL3-8B 44.6% 8-shot Apr 2025 Official leaderboard
5 Qwen2.5-VL-7B 42.7% Chain of thought Jan 2025 Official leaderboard
6 Janus-Pro-7B 40.3% Chain of thought Jan 2025 Official leaderboard
7 Gemma 3 4B 35.1% 8-shot Mar 2025 Official leaderboard
8 LLaVA-v1.5-13B 29.7% 2-shot Oct 2023 Official leaderboard
9 llava-v1.5-7b 24.1% 2-shot Oct 2023 Official leaderboard

Rankings by part

We read 9 parts from the benchmark’s paper, each ranking models on one part of the test, such as Processing correlation, Morphology analysis and Structure analysis. Llama 4 Maverick leads every one of them. Arrows show how far a model moved from its overall rank among the models we track.

Processing correlation

9 models we track have a score here.

  1. 1 Llama 4 Maverick 88.2%
  2. 2 InternVL3-38B 75.2%
  3. 3 Qwen2.5-VL-7B ↑2 71.2%
  4. 4 Qwen2.5-VL-32B ↓1 69.3%
  5. 5 Gemma 3 4B ↑2 64.7%
  6. 6 InternVL3-8B ↓2 62.1%
  7. 7 Janus-Pro-7B ↓1 55.6%
  8. 8 LLaVA-v1.5-13B 43.1%
  9. 9 llava-v1.5-7b 26.1%

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
Llama 4 Maverick
Chain of th… Zero-shot 4-shot
68.1%
InternVL3-38B
Zero-shot
52.2%
Qwen2.5-VL-32B
2-shot Zero-shot
50.1%
InternVL3-8B
Zero-shot 4-shot
44.6%
Qwen2.5-VL-7B
16-shot
42.7%
Janus-Pro-7B
8-shot Zero-shot 16-shot
40.3%
Gemma 3 4B
4-shot
35.1%
LLaVA-v1.5-13B
8-shot Chain of th… Zero-shot
29.7%
llava-v1.5-7b
8-shot Chain of th… 4-shot
24.1%

Source: The MatCha paper

ModelEffortHarnessScoreCost per taskOutput tokens
Llama 4 Maverick 16-shot 68.1%
Llama 4 Maverick 2-shot 66.9%
Llama 4 Maverick 8-shot 66.6%
Llama 4 Maverick 4-shot 65.9%
Llama 4 Maverick Zero-shot 54.8%
InternVL3-38B 2-shot 52.2%
InternVL3-38B 4-shot 51.7%
InternVL3-38B Chain of thought 50.9%
InternVL3-38B 8-shot 50.4%
InternVL3-38B 16-shot 50.3%
Qwen2.5-VL-32B Chain of thought 50.1%
Qwen2.5-VL-32B Zero-shot 49.9%
InternVL3-38B Zero-shot 48.7%
Llama 4 Maverick Chain of thought 45.7%
Qwen2.5-VL-32B 4-shot 45.3%
Qwen2.5-VL-32B 16-shot 45.1%
InternVL3-8B 8-shot 44.6%
InternVL3-8B 2-shot 44.3%
InternVL3-8B 16-shot 44.1%
Qwen2.5-VL-32B 8-shot 44.1%
Qwen2.5-VL-32B 2-shot 43.5%
Qwen2.5-VL-7B Chain of thought 42.7%
InternVL3-8B 4-shot 42.1%
InternVL3-8B Chain of thought 41.6%
Qwen2.5-VL-7B 4-shot 41.3%
Qwen2.5-VL-7B Zero-shot 41%
Qwen2.5-VL-7B 2-shot 40.6%
Janus-Pro-7B Chain of thought 40.3%
Janus-Pro-7B 16-shot 40.2%
Qwen2.5-VL-7B 8-shot 39.9%
Qwen2.5-VL-7B 16-shot 38.9%
Janus-Pro-7B 2-shot 36.5%
InternVL3-8B Zero-shot 36.4%
Gemma 3 4B 8-shot 35.1%
Janus-Pro-7B 4-shot 34.9%
Gemma 3 4B 16-shot 34.8%
Janus-Pro-7B Zero-shot 34.5%
Gemma 3 4B Zero-shot 34.3%
Gemma 3 4B Chain of thought 34.1%
Gemma 3 4B 2-shot 32.5%
Gemma 3 4B 4-shot 32%
LLaVA-v1.5-13B 2-shot 29.7%
LLaVA-v1.5-13B Zero-shot 27.9%
llava-v1.5-7b 2-shot 24.1%
LLaVA-v1.5-13B 4-shot 20.7%
llava-v1.5-7b Zero-shot 20.5%
llava-v1.5-7b 4-shot 19.9%
LLaVA-v1.5-13B Chain of thought 15.2%
llava-v1.5-7b Chain of thought 13.9%
LLaVA-v1.5-13B 8-shot 0%
LLaVA-v1.5-13B 16-shot 0%
llava-v1.5-7b 8-shot 0%
llava-v1.5-7b 16-shot 0%
Janus-Pro-7B 8-shot 0%

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
Llama 4 Maverick
Chain of th… Zero-shot 4-shot
68.1%
InternVL3-38B
Zero-shot
52.2%
Qwen2.5-VL-32B
2-shot Zero-shot
50.1%
InternVL3-8B
Zero-shot 4-shot
44.6%
Qwen2.5-VL-7B
16-shot
42.7%
Janus-Pro-7B
8-shot Zero-shot 16-shot
40.3%
Gemma 3 4B
4-shot
35.1%
LLaVA-v1.5-13B
8-shot Chain of th… Zero-shot
29.7%
llava-v1.5-7b
8-shot Chain of th… 4-shot
24.1%

Source: The MatCha paper

ModelEffortHarnessScoreCost per taskOutput tokens
Llama 4 Maverick 16-shot 68.1%
Llama 4 Maverick 2-shot 66.9%
Llama 4 Maverick 8-shot 66.6%
Llama 4 Maverick 4-shot 65.9%
Llama 4 Maverick Zero-shot 54.8%
InternVL3-38B 2-shot 52.2%
InternVL3-38B 4-shot 51.7%
InternVL3-38B Chain of thought 50.9%
InternVL3-38B 8-shot 50.4%
InternVL3-38B 16-shot 50.3%
Qwen2.5-VL-32B Chain of thought 50.1%
Qwen2.5-VL-32B Zero-shot 49.9%
InternVL3-38B Zero-shot 48.7%
Llama 4 Maverick Chain of thought 45.7%
Qwen2.5-VL-32B 4-shot 45.3%
Qwen2.5-VL-32B 16-shot 45.1%
InternVL3-8B 8-shot 44.6%
InternVL3-8B 2-shot 44.3%
InternVL3-8B 16-shot 44.1%
Qwen2.5-VL-32B 8-shot 44.1%
Qwen2.5-VL-32B 2-shot 43.5%
Qwen2.5-VL-7B Chain of thought 42.7%
InternVL3-8B 4-shot 42.1%
InternVL3-8B Chain of thought 41.6%
Qwen2.5-VL-7B 4-shot 41.3%
Qwen2.5-VL-7B Zero-shot 41%
Qwen2.5-VL-7B 2-shot 40.6%
Janus-Pro-7B Chain of thought 40.3%
Janus-Pro-7B 16-shot 40.2%
Qwen2.5-VL-7B 8-shot 39.9%
Qwen2.5-VL-7B 16-shot 38.9%
Janus-Pro-7B 2-shot 36.5%
InternVL3-8B Zero-shot 36.4%
Gemma 3 4B 8-shot 35.1%
Janus-Pro-7B 4-shot 34.9%
Gemma 3 4B 16-shot 34.8%
Janus-Pro-7B Zero-shot 34.5%
Gemma 3 4B Zero-shot 34.3%
Gemma 3 4B Chain of thought 34.1%
Gemma 3 4B 2-shot 32.5%
Gemma 3 4B 4-shot 32%
LLaVA-v1.5-13B 2-shot 29.7%
LLaVA-v1.5-13B Zero-shot 27.9%
llava-v1.5-7b 2-shot 24.1%
LLaVA-v1.5-13B 4-shot 20.7%
llava-v1.5-7b Zero-shot 20.5%
llava-v1.5-7b 4-shot 19.9%
LLaVA-v1.5-13B Chain of thought 15.2%
llava-v1.5-7b Chain of thought 13.9%
LLaVA-v1.5-13B 8-shot 0%
LLaVA-v1.5-13B 16-shot 0%
llava-v1.5-7b 8-shot 0%
llava-v1.5-7b 16-shot 0%
Janus-Pro-7B 8-shot 0%

Number one over time

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

Meta
Benchmark paper Not tested since Sep 2025

Source: The MatCha 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 Best so far
0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Llama 4 Maverick Qwen2.5-VL-32B Qwen2.5-VL-7B LLaVA-v1.5-13B 0% 20% 40% 60% 80% 100% 2024 2025 Near the ceiling (85%) Llama 4 Maverick Qwen2.5-VL-32B

Source: Official MatCha 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

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 19 months ago, in April 2025, for InternVL3-38B. 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 benchmark’s paper (15 entries).
  • Pass Models scored 9 models we track, plus 6 other entries on its leaderboard.
  • Warning Newest score The newest came 19 months ago, in April 2025, for InternVL3-38B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Llama 4 Maverick (#153).
  • Pass Headroom The best is 68.1%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 54 rows for models we track and 6 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