physics benchmark
PHYBench
PHYBench is a benchmark of real-world physics problems spanning mechanics, electromagnetism, thermodynamics, optics, and modern physics, designed to evaluate physical perception and multi-step quantitative reasoning in large language models.
Updated Aug 7, 2026
Published models1
Registry coverage1
MetricScore
EvidenceB
PHYBench leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | TE | 77.4% | 100.0% | 1 | C |
PHYBench highlights
The top published results on this benchmark's own scale.
What is PHYBench?
Definition and scoring fields from the benchmark registry.
PHYBench is a benchmark of real-world physics problems spanning mechanics, electromagnetism, thermodynamics, optics, and modern physics, designed to evaluate physical perception and multi-step quantitative reasoning in large language models.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- PHYBench
- Modality
- text
- Primary category
- physics
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- phybench|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about PHYBench.
Which model scores highest on PHYBench?
Hy3 is currently ranked first with 77.4%.
What does PHYBench measure?
PHYBench is a benchmark of real-world physics problems spanning mechanics, electromagnetism, thermodynamics, optics, and modern physics, designed to evaluate physical perception and multi-step quantitative reasoning in large language models.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.