Mistral Large 4 Release - Cognition FrontierCode 2.0: leaderboard
Mistral Large 4 open-weight release post (2026-10-29): Cognition's private FrontierCode 2.0 benchmark, which assesses end-to-end code quality and whether a repository maintainer would merge the agent's change, as reported by Mistral for six models. Harness, attempts and reasoning effort not stated; separate from Artificial Analysis's FrontierCode 2.0 run. Score (%).
Metric: Score (%). Source: mistral.ai. Saturation forecast: Around July 2028. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5.3 | 17.8 | #44 |
| 2 | Kimi K3 | 17.3 | #35 |
| 3 | GLM-5.3 Flash | 14.2 | #77 |
| 4 | Beam 1.0 | 2.9 | #102 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mistral-large-4-release-cognition-frontiercode-2-0 · How It Works · Data refreshed daily, snapshot 2026-11-05.