Vendor
6 model versions evaluated ·27 runs in the corpus · rubric v15 · Google site ↗
Model versions
| Model | Meta-score | Appearances | Best at | Last seen |
|---|---|---|---|---|
| Google Gemini Flash 3 (preview) | 61.8 | 6 | Scout Agent | 2026-07-10 |
| Google Gemini Flash 3.5 | 59.9 | 8 | Develop Agent | 2026-07-16 |
| Google Gemini Flash 3.6 | 72.3 | 5 | Code-Review Agent | 2026-07-24 |
| Google Gemini Flash Lite 3.1 | 74.3 | 4 | Sprint-Review Agent | 2026-07-19 |
| Google Gemini Flash Lite 3.5 | 71.0 | 1 | Develop Agent | 2026-07-21 |
| Google Gemini Pro 3.1 (preview) | 58.0 | 2 | Test Agent | 2026-07-03 |
Runs (primary model)
Every run on Favur Evals is scored by the same deterministic engine, on the same Statements of Work. The benchmark is self-funded — no vendor sponsorship, credits, or grants.