Glossary
SimpleQA Verified
SimpleQA Verified, on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), asks short factual questions with a single right answer, reviewed to remove answer-key errors. The score is the share of correct answers.
What it measures
SimpleQA Verified, on the Epoch AI leaderboard (CC BY 4.0, latest version we checked), asks short factual questions with a single right answer, reviewed to remove answer-key errors. The score is the share of correct answers.
When it matters
It matters when a made-up fact is costly, as in a support reply or research that goes to a client. It measures factual recall. A model that reasons well can still do poorly here.
How to read it
The score is the share of short facts answered correctly, on the Epoch AI leaderboard, in the latest version we checked. With a low score, I would pair the model with search or documents, even if it reasons well.
Who leads
- Claude Fable 5.2 Anthropic94%
- Gemini 4 Argon Google76%
- GPT-6 Astra OpenAI75.6%
- GPT-6.1 Sol OpenAI73.9%
- Gemini 3.1 Pro Google73.5%
- Claude Opus 5.5 Anthropic72.2%
- Claude Fable 5.1 Anthropic70.8%
- Claude Fable 5 Anthropic70.7%
- Gemini 3.8 Flash Google69.7%
- GPT-5.6 Sol OpenAI69.7%
Source: Epoch AI, latest version we checked.