SWE-bench Pro
Full rankings
3 models tested · sorted by score · includes 3 verified scores
| # | Model | Score |
|---|---|---|
| 1 | 77.8 | |
| 2 | 64.3 | |
| 3 | 53.4 |
Score distribution
Where models cluster
Correlated benchmarks
Pearson r · original research
Frequently asked
About SWE-bench Pro
What does SWE-bench Pro measure?
SWE-bench Pro is a knowledge benchmark in the BenchGecko catalog. 1 AI models have been tested on it. Scores range from 77.8 to 77.8 out of 100.
Which model leads on SWE-bench Pro?
Claude Mythos Preview from Anthropic leads SWE-bench Pro with a score of 77.8. The median score across 1 tested models is 77.8.
Is SWE-bench Pro saturated?
No · the top score is 77.8 out of 100 (78%). There is still meaningful room for improvement on SWE-bench Pro.
What makes SWE-bench Pro distinctive?
SWE-bench Pro is a knowledge benchmark with limited overlap to the rest of the catalog · it measures capabilities that are not well-covered by other benchmarks we track.
How often is SWE-bench Pro data refreshed?
BenchGecko pulls updates daily. New model scores on SWE-bench Pro appear as soon as they are published by Epoch AI or the model provider.
- Category
- Knowledge
- Max score
- 100
- Models
- 3
- Updated
- 2026-04-07
More knowledge benchmarks
Same category · related evaluations