FrontierSWE
FrontierSWE results reported for 19 models across 2 versions. Updated 2026-10-08.
Report an issueVersions
2 versions · Metric: FrontierSWE Score
| Version | Released | Models | Results |
|---|---|---|---|
| FrontierSWE 2 — October 2026 evaluationConfiguration: Proximus harness, 5 trials, 20-hour budget | October 7, 2026 | 18 | 18 |
| FrontierSWE 2 — September 2026 evaluationLatest | September 30, 2026 | 1 | 1 |
Recently reported results: FrontierSWE 2 — September 2026 evaluation
Latest version: FrontierSWE 2 — September 2026 evaluation
| Model | Score | Source |
|---|---|---|
| Gemini 4 Argon (highest thinking settings) | 55.0% | deepmind.google for Gemini 4 Argon (highest thinking settings) on FrontierSWE 2 — September 2026 evaluation (opens in a new tab)FrontierSWE ScoreReported Evaluation methodologyGoogle DeepMindSelf-reportedProximalRecord for Gemini 4 Argon on FrontierSWE 2 — September 2026 evaluation |