@dabit3: FrontierCode is the first eval to measure the metric that matters most in real software engineering: would you actually…
Summary
FrontierCode is a new coding evaluation benchmark that measures code mergeability, claiming 81% fewer misclassification errors than SWE-Bench Pro. Tasks were crafted by maintainers of open-source projects like Celery, uppy, and Mattermost.
View Cached Full Text
Cached at: 06/08/26, 11:28 PM
FrontierCode is the first eval to measure the metric that matters most in real software engineering: would you actually merge this code?
The result is 81% fewer misclassification errors than SWE-Bench Pro, the most accurate model ranking available today.
Every task was crafted by the actual maintainers of repos like Celery, uppy, and Mattermost. I also encourage you to dive into the blog post, it’s beautiful and interactive!
Cognition (@cognition): Introducing FrontierCode: a coding eval that raises the bar for difficulty & quality. Each task took 40+ hrs of work by leading open-source maintainers.
Models write sloppy code that works but isn’t maintainable. Our eval is first to measure: would you actually merge this code?
Similar Articles
@dabit3: FrontierCode measures "would a maintainer merge this PR?" Essentially intelligence + codebase fluency + judgement, whic…
Cognition introduces FrontierCode 2.0, a coding benchmark that evaluates whether a model's pull request would actually be merged by a maintainer — testing intelligence, codebase fluency, and judgement beyond simply meeting a spec.
@denizbirlikci: To understand why we built FrontierCode, read @METR_Evals's blog post on why "many SWE-bench-passing PRs would not be m…
Cognition announces FrontierCode, a new coding evaluation benchmark that goes beyond unit tests to measure code quality, scope, test correctness, and human reviewer approval, addressing the issue of agents writing sloppy code that passes tests but is not maintainable.
FrontierCode
FrontierCode is a new benchmark from Cognition AI that measures AI models' ability to write high-quality, maintainable code by evaluating mergeability. Results show even top models like Claude Opus 4.8 score only 13.4% on the hardest subset, highlighting a significant gap in code quality.
@swyx: It's finally out!!! @METR_Evals found that more than half of SWEBench results is unmergeable slop. FrontierCode represe…
FrontierCode is a new coding benchmark from METR and Cognition that evaluates AI models on code maintainability and quality, revealing that many models produce unmergeable code. It includes over 1000 hours of work and shows that even top models struggle, with Opus 4.8 achieving only 13.8% on the hardest tier.
FrontierCode: a coding eval that raises the bar for difficulty & quality.
FrontierCode is a new coding evaluation benchmark designed to increase difficulty and quality standards for AI code generation.