Epoch AI released MirrorCode, a long-horizon programming benchmark focused on measuring an AI's ability to complete large, real codebase refactors end-to-end. The evaluation subject this time: Claude Opus 4.7 completes a 60,000-line Pkl configuration-language rewrite in a single task, taking about 7 days of wall-clock time, with full test coverage passing.
MirrorCode's design philosophy is different from SWE-Bench: it doesn't break a fix into "edit a few lines" units, but presents a "rewrite 60,000 lines while keeping all behavior consistent" megatask. That makes it a better mirror of real engineering work, and a more honest measure of an AI's long-horizon programming capability.
The result: Opus 4.7 successfully completes the rewrite with no human intervention, and the code quality and test pass rate are on par with a senior engineer's delivery. Epoch AI says this represents a doubling of long-horizon programming capability in the past year, and the inflection came from three things: tool-use stability, context consistency, and test-driven self-correction.
The most important takeaway: MirrorCode is not a "leaderboard" — it's a "stress test." It pushes AI coding to the boundary of real engineering, exposing the gaps that "single-function completion" benchmarks hide. For developer-tool vendors, this means the next competitive focus is not "how many lines of code" but "how complex a real refactor can you take on."