Eval / benchmark
Run a test match with your own settings to find the speed/quality trade-off. Results are recorded per home and per model/effort. This never changes the live config.
1 · Pick an open shift
2 · Settings for this run
History
Run-to-run drift on the same shift is expected — bookings move the pool, so pool keeps comparisons honest.
| When | Home | Model · effort | judged | pool | total | judge | top 3 |
|---|---|---|---|---|---|---|---|
| No runs yet. | |||||||