FrontierSWE
Ultra-long-horizon coding benchmark for frontier agents
Community-submitted · content unverified. This organization has not verified ownership or the information in this profile.
About
FrontierSWE is Proximal's ultra-long-horizon coding benchmark for evaluating frontier models on hard, multi-hour software engineering tasks. FrontierSWE v2 expands the suite with longer budgets and broader model comparisons published on Proximal's site.
- Website
- proximal.ai
- Added via
- web
- Ownership
- unclaimed