Scores updated · 13 tests both models report · How we compare
Every test, side by side
All 13 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Other results13 tests
- NarrativeQADeepSeek-LLM-67B-Chat (V1) by 4711.158.1+47
- naturalquestions_closedbookDeepSeek-LLM-67B-Chat (V1) by 38.42.841.2+38.4
- AIR-Bench 2024Claude 3 Sonnet by 34.284.750.5+34.2
- HarmBenchClaude 3 Sonnet by 30.995.864.9+30.9
- MATHDeepSeek-LLM-67B-Chat (V1) by 18.443.161.5+18.4
- MMLUClaude 3 Sonnet by 14.278.364+14.2
- GSM8KClaude 3 Sonnet by 12.892.379.5+12.8
- bbqClaude 3 Sonnet by 3.89086.2+3.8
- OpenBookQAClaude 3 Sonnet by 3.891.888+3.8
- simple_safety_testsClaude 3 Sonnet by 3.210096.8+3.2
- XSTestDeepSeek-LLM-67B-Chat (V1) by 3.185.888.9+3.1
- AA Intelligencetie5.95.3tie
- anthropic_red_teamtie99.899.4tie
Questions people ask
How do you compare the two?
We use the 13 benchmark tests both models have published scores on. None of them is in the eight capability areas we count, so this page lists them without a verdict. Each score is the one shown on the model's own page.