math benchmark
OmniMath Leaderboard
OmniMath is a mathematics benchmark for large language models containing 4,428 competition-level problems with rigorous human annotation across over 33 sub-domains and more than 10 distinct difficulty levels.
Updated Sep 5, 2026
Models2
Model coverage2
MetricScore
EvidenceB
OmniMath Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelMI | Score81.90% | Percentile100.00% | Participants2 | EvidenceC | Evaluated |
| Rank02 | ModelMI | Score76.60% | Percentile0.00% | Participants2 | EvidenceC | Evaluated |
OmniMath Highlights
The leading models and scores on this benchmark.
OmniMath Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for OmniMath
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is OmniMath?
What OmniMath measures and how its scores work.
OmniMath is a universal Olympiad-level mathematics benchmark for large language models.
It reports a Score metric expressed as a ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- OmniMath
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about OmniMath.
Which model scores highest on OmniMath?
Phi 4 Reasoning Plus is currently ranked first with 81.90%.
What are the top three models on OmniMath?
The current leaders are Phi 4 Reasoning Plus (81.90%) and Phi 4 Reasoning (76.60%).
Which OmniMath model has the lowest official input price?
Phi 4 Reasoning Plus has the lowest matched official input price at $0.13 input / $0.50 output per 1M tokens.
Which models are fastest among OmniMath results?
No matched runtime record is currently available.
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.