math benchmark
LiveMathematicianBench
LiveMathematicianBench evaluates research-level mathematical reasoning on continuously refreshed, contamination-resistant problems.
Updated Aug 11, 2026
Models2
Model coverage2
MetricScore
EvidenceB
LiveMathematicianBench Ranking
Higher score ranks better on this benchmark.
| 01 | BY | 27.7% | 100.0% | 2 | C | |
| 02 | BY | 20.9% | 0.0% | 2 | C |
LiveMathematicianBench Score Distribution
A closer view of the leading scores on this benchmark.
LiveMathematicianBench Highlights
The leading models and scores on this benchmark.
What is LiveMathematicianBench?
What LiveMathematicianBench measures and how its scores work.
LiveMathematicianBench evaluates research-level mathematical reasoning on continuously refreshed, contamination-resistant problems.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- LiveMathematicianBench
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- livemathematicianbench|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about LiveMathematicianBench.
Which model scores highest on LiveMathematicianBench?
Seed 2.1 Turbo is currently ranked first with 27.7%.
What does LiveMathematicianBench measure?
LiveMathematicianBench evaluates research-level mathematical reasoning on continuously refreshed, contamination-resistant problems.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.