Coding and terminal agents / Benchmark leaderboard
DeepSWE 1.1
113 original long-horizon repository tasks across five programming languages.
No published company result yet.
This benchmark is in the research directory. Its task package, adapter and grading protocol need qualification before a hosted run can be offered.
Arrange an evaluation for your company →What this benchmark measures
Metrics
Behavioral verifier success.
Execution requirements
Harbor with a separate pristine verifier environment.
Scope and limitations
Pin 1.1 and verifier runtime; the reference patch is not the grading criterion. Do not confuse this benchmark with similarly named models.
Evaluation availability
Catalog entry. Request a managed evaluation to qualify your agent interface and the benchmark’s native grading requirements.
Sources and company fit
- Original benchmark ↗Newer specialist
- Hugging Face · datacurve/deep-swe ↗Gated · Selected newer long-horizon coding family; pin 1.1 versus prior release.
Companies whose products may fit
Research recommendations based on product capabilities. These companies have not necessarily run this benchmark or integrated with Blobfish.
Compatibility notes for each company
All Hands / OpenHands: Capability-aligned; adapter/access to qualify
Augment Code: Capability-aligned; adapter/access to qualify
Cline: Capability-aligned; adapter/access to qualify
Cognition / Devin: Capability-aligned; adapter/access to qualify
Cursor: Capability-aligned; adapter/access to qualify
Factory: Capability-aligned; adapter/access to qualify
GitHub Copilot: Capability-aligned; adapter/access to qualify
JetBrains: Capability-aligned; adapter/access to qualify
Replit: Capability-aligned; adapter/access to qualify
Sourcegraph: Capability-aligned; adapter/access to qualify