agents benchmark
SWE-Atlas
SWE-Atlas is a software engineering benchmark focused on debugging, evaluating a model's ability to localize and fix bugs in real-world codebases.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
SWE-Atlas leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | BY | 35.2% | 100.0% | 2 | C | |
| 2 | BY | 30.6% | 0.0% | 2 | C |
Score distribution
Top published rows on the benchmark's original scale.
SWE-Atlas highlights
The top published results on this benchmark's own scale.
What is SWE-Atlas?
Definition and scoring fields from the benchmark registry.
SWE-Atlas is a software engineering benchmark focused on debugging, evaluating a model's ability to localize and fix bugs in real-world codebases.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- SWE-Atlas
- Modality
- text
- Primary category
- agents
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- swe-atlas|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about SWE-Atlas.
Which model scores highest on SWE-Atlas?
Seed 2.1 Pro is currently ranked first with 35.2%.
What does SWE-Atlas measure?
SWE-Atlas is a software engineering benchmark focused on debugging, evaluating a model's ability to localize and fix bugs in real-world codebases.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.