Benchmarks

SWE-bench Science

Benchmark · From Fudan University · 1 item since Aug 2026

Repository-level coding tasks across 98 scientific-software repositories in 20 domains.

Items

  1. Aug 15

    Fudan's SWE-bench Science stumps top agents

    A repository-level benchmark of 119 scientific-software tasks across 98 repositories and 20 domains on which the strongest agent, Claude Code with Opus 5 at max effort, scores below 50% pass@1, with scientific guidance helping only when well aligned.