reasoning benchmark
SimpleQA
SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.
Updated Aug 7, 2026
Published models46
Registry coverage46
MetricScore
EvidenceB
SimpleQA leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
Score distribution
Top published rows on the benchmark's original scale.
SimpleQA highlights
The top published results on this benchmark's own scale.
What is SimpleQA?
Definition and scoring fields from the benchmark registry.
SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- SimpleQA
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- simpleqa|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about SimpleQA.
Which model scores highest on SimpleQA?
DeepSeek-V3.2-Exp is currently ranked first with 97.1%.
What does SimpleQA measure?
SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
46 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is marked as eligible for the current LLMBoard capability methodology.