reasoning benchmark
OmniScience
OmniScience is a broad scientific knowledge and reasoning benchmark that measures both answer accuracy and non-hallucination (calibrated abstention) across science domains.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
OmniScience leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | NV | 78.7% | 100.0% | 2 | C | |
| 2 | XA | 52.0% | 0.0% | 2 | B |
Score distribution
Top published rows on the benchmark's original scale.
OmniScience highlights
The top published results on this benchmark's own scale.
What is OmniScience?
Definition and scoring fields from the benchmark registry.
OmniScience is a broad scientific knowledge and reasoning benchmark that measures both answer accuracy and non-hallucination (calibrated abstention) across science domains.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- OmniScience
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- omniscience|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about OmniScience.
Which model scores highest on OmniScience?
Nemotron 3 Ultra (550B A55B) is currently ranked first with 78.7%.
What does OmniScience measure?
OmniScience is a broad scientific knowledge and reasoning benchmark that measures both answer accuracy and non-hallucination (calibrated abstention) across science domains.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 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.