OmniScience (non-hallucination rate)
Progress Over Time
Interactive timeline showing model performance evolution on OmniScience (non-hallucination rate)
OmniScience (non-hallucination rate) Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 |  Institute of Foundation Models | 375B | — | — | ||
| 2 | Institute of Foundation Models | 32B | — | — | ||
| 3 | Institute of Foundation Models | 36B | — | — | ||
| 4 |  xAI | — | 500K | $2.00 / $6.00 |
What is OmniScience (non-hallucination rate)?
OmniScience variant that reports the non-hallucination rate, defined as one minus the hallucination rate.
OmniScience (non-hallucination rate) is a text benchmark evaluating models on knowledge, reasoning, and science tasks. LLM Stats tracks 4 models on this benchmark, scored on a 0–1 scale. The current average is 0.6, with the leader at 0.7.
Compare leaders on the best AI for knowledge, best AI for reasoning and best AI for science leaderboards.
Current leaders
K2 Horizon 375B-A23B from Institute of Foundation Models currently leads the OmniScience (non-hallucination rate) leaderboard with a score of 0.747 across 4 evaluated AI models.
FAQ
Common questions about the OmniScience (non-hallucination rate) benchmark and leaderboard.