knowledge benchmark
OmniScience (non-hallucination rate)
OmniScience variant that reports the non-hallucination rate, defined as one minus the hallucination rate.
Updated Aug 11, 2026
Models1
Model coverage1
MetricScore
EvidenceB
OmniScience (non-hallucination rate) Ranking
Higher score ranks better on this benchmark.
| 1 | XA | 46.0% | 100.0% | 1 | B |
OmniScience (non-hallucination rate) Highlights
The leading models and scores on this benchmark.
What is OmniScience (non-hallucination rate)?
What OmniScience (non-hallucination rate) measures and how its scores work.
OmniScience variant that reports the non-hallucination rate, defined as one minus the hallucination rate.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- OmniScience (non-hallucination rate)
- Modality
- text
- Primary category
- knowledge
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- omniscience-non-hallucination-rate|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about OmniScience (non-hallucination rate).
Which model scores highest on OmniScience (non-hallucination rate)?
Grok 4.5 is currently ranked first with 46.0%.
What does OmniScience (non-hallucination rate) measure?
OmniScience variant that reports the non-hallucination rate, defined as one minus the hallucination rate.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 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.