FrontierCode
Coding benchmark (Cognition's own bench per S8)
v1.1codingactive
Definition
- Version
- v1.1
- Category
- coding
- Maintainer
- Cognition (owns Windsurf)
- Status
- active
- Home page
- not in the notes
Sources
- Claude Opus 5.5: Fable-Class Work, Cheaper to Run llm-stats.com https://llm-stats.com/blog/research/claude-opus-5-5-launch
- Cognition SWE-2 : Modèle, Harnais et les chiffres qui se cachent derrière orcarouter.ai https://www.orcarouter.ai/fr/blog/swe-2
Scores at a Glance
Score in %
Higher is better for this unit.
0255075100
Where this comes from
Each bar is one model, with its most independent recorded score. Hatched bars are vendor-reported; the word beside each bar says who ran it. Settings and dates differ between rows; the Scores table below has every result.
Scores
| Model | Score | Who ran it | Setting | View or population | Date | Source |
|---|---|---|---|---|---|---|
 Claude Opus 5.5 | 54.4 % | vendor-reported | – | – | – | llm-stats.com |
| SWE-2 | 50 % | vendor-reported | – | – | – | orcarouter.ai |
Where this comes from
Scores from different settings, harnesses or dates are not directly comparable; the row says which. Sorted by score, highest first, as a reading aid, not a ranking.
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "FrontierCode", https://aurion.semai.id/benchmarks/frontiercode-v1-1/, records of October 2026.