Artificial Analysis launches FrontierCode 1.1 leaderboard for Cognition's coding benchmark
Article timeline
The articles in this story. Times are ET.
Artificial Analysis@ArtificialAnlysOfficialArtificial Analysis highlights FrontierCode, a Cognition coding benchmarkAIArtificial Analysis links to full results for FrontierCode, a new evaluation introduced by Cognition. The post includes no scores, models, or methodology details beyond the links to the Artificial Analysis evaluation page and Cognition's blog post.
- Artificial Analysis@ArtificialAnlysOfficialArtificial Analysis launches FrontierCode 1.1 leaderboard for maintainer-ready coding
AIArtificial Analysis launches a leaderboard for FrontierCode 1.1, a private coding benchmark from Cognition that grades whether changes would be merged by repository maintainers. Claude Opus 5.5 (max with fallback) leads at 52%, and nine of 100 tasks went unsolved by every model tested. GPT-6.1 Sol (max) scores 49% at $0.81 per task, matching Claude Fable 5.1 (max with fallback) at less than a tenth of the cost.
- Artificial Analysis ArticlesOfficialArtificial Analysis launches FrontierCode benchmark for maintainer-merge coding tasks
AIArtificial Analysis launched FrontierCode 1.1, a coding benchmark built by Cognition with maintainers of 36 open source repositories. Claude Opus 5.5 (max with fallback) leads with a 52% average score and a 58% pass rate, and leading models still fail more than 40% of the time. The benchmark also penalizes excessive changes, out-of-scope file edits, and unhelpful tests, not just failing tests.