FrontierCode 1.1 Main
Cognition's FrontierCode 1.1 Main evaluation measures whether coding agents produce mergeable changes on the 100 hardest tasks from its 150-task Extended set. It reports a mean@5 composite of held-out test performance and weighted code quality; Main remains distinct from the full Extended variant.
Models ranked
17
tracked on this benchmark
Score band
54.6 – 31.8
best → lowest tracked
Snapshot trend
-7.39
Jul 24 → Oct 5 · 15 models
Leaderboard
Tracked models ranked by Mean@5 score (higher is better).
Configuration: claude-opus-5-5 (Medium)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: claude-code. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: Medium effort
Notes: Source: p.150, Figure 8.3. Medium is the reported Opus 5 setting; preserve it and keep Main separate from Extended.
Configuration: gpt-6-astra (Max)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: codex. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: claude-sonnet-5-5 (Extra High)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: claude-code. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: claude-fable-5-1 (Medium)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: claude-code. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: gpt-6.1-sol (Medium)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: codex. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: swe-2 (Max)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: devin. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: gpt-6-sol (Max)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: codex. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: grok-4.6 (High)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: grok-build. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: grok-4.7 (High)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: grok-build. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: kimi-k3 (None)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: mini-swe-agent. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: gpt-6-luna (Max)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: codex. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: gemini-3.8-flash (Medium)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: chisel. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: glm-5.3 (Max)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: chisel. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: gemini-3.6-flash (Medium)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: chisel. Highest-scoring published effort selected; timestamp is retrieval date.
Configuration: glm-5.3-flash (Max)
Notes: Cognition FrontierCode 1.1 Main (100 tasks) mean@5 composite ('new_score'), verified in https://cognition.com/data/frontiercode-leaderboard/data.json (the data file behind https://cognition.com/frontiercode), retrieved 2026-10-05. Agent harness: chisel. Highest-scoring published effort selected; timestamp is retrieval date.
How to read this benchmark
This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.
Trust this score when
- There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
- The model list covers the same version family you can actually deploy today.
- Top candidates overlap with your required routing and feature requirements.
Be cautious when
- There is only one benchmark snapshot or the dataset appears stale.
- The benchmark metric direction is opposite of your decision objective.
- The score difference between options is narrow and likely within implementation variance.
Related benchmarks
Last reviewed: Jul 26, 2026