batchwatch › OpenAI
OpenAI publishes a completion window of 24 hours and nothing more precise. Below is what we have actually measured, model by model, with the confidence each figure carries.
30-day window. A dash means we have measurements but fewer than 2 - not enough to quote a percentile, so we do not.
| Model | p50 | p90 | p95 | Jobs | Sources | Confidence | Last measured |
|---|---|---|---|---|---|---|---|
| gpt-5.6-sol | 78s | 2 min | 3 min | 10 | 1 | Confidence: Very low | 2026-08-25 |
| gpt-5.6-luna | - | - | - | 6 | 1 | Confidence: Very low | 2026-08-25 |
| gpt-5-nano | - | - | - | 5 | 1 | Confidence: Very low | 2026-08-25 |
No key required for coverage.
curl "https://batchwatch.dev/v1/coverage"
Per-model curves live at
/v1/curve?provider=openai&model=....
The front page has the decision endpoints.