Benchmarks · SWE-bench Full
Run planned
Full run (2 294 instances) is planned once the Lite model matrix completes. Estimated cost ~$20-40, ~60 hours wall-clock with sharding.
Back to Benchmarks overviewBenchmarks · SWE-bench Full
Full run (2 294 instances) is planned once the Lite model matrix completes. Estimated cost ~$20-40, ~60 hours wall-clock with sharding.
Back to Benchmarks overview