Benchmark Registry

SimpleQA Verified

SimpleQA Verified results reported for 59 models across 1 version. Updated 2026-10-07.

Report an issue

Versions

1 version, plus 1 variant or configuration · Metric: Accuracy

VersionReleasedModelsResults
SimpleQA Verified OriginalLatest
1 variant of Original
VariantConfigurationReleasedModelsResults
SimpleQA Verified Original — Epoch AI harnessEpoch AI harnessSeptember 9, 20255764
September 9, 202559

Recently reported results: SimpleQA Verified Original

Latest version: SimpleQA Verified Original

ModelScoreSource
MAI-Thinking-131.0%microsoft.aiAccuracyReported Technical reportMicrosoft AISelf-reportedMicrosoftRecord
Inkling-Small (effort=0.99)20.6%thinkingmachines.aiAccuracyReported Launch postThinking Machines LabSelf-reportedRecord
Inkling (effort=0.99)43.9%thinkingmachines.aiAccuracyReported Launch postThinking Machines LabSelf-reportedRecord
DeepSeek-V4-Pro (max)57.9%huggingface.coAccuracyReported Model cardDeepSeekSelf-reportedDeepSeek AIRecord
DeepSeek-V4-Pro (high)46.2%huggingface.coAccuracyReported Model cardDeepSeekSelf-reportedDeepSeek AIRecord