FrontierSWE (Dominance)
Code · 2026-04-16
FrontierSWE V1 is an ultra-long-horizon software-engineering benchmark. Its dominance score is the win rate against a random opponent on a random task, as defined in the publisher's V1 archive; it is distinct from later FrontierSWE mean@5 measurements.
Top models (higher is better)
| Model | Score |
|---|---|
| Kimi K3 | 81.2 |
| GLM-5.3 | 78.1 |
| GPT-5.6 Sol | 71.3 |
| GLM-5.2 | 67.3 |
| Opus 4.8 | 66.7 |
| GPT-5.5 | 64.9 |