OpenAI
davinci-002
Made by OpenAI. Release date not recorded. Accessibility not recorded. Cheapest standard rate we hold: $2 per million output tokens, at long context.
What it costs
| Tier | Kind | In | Out | Cached in | From |
|---|---|---|---|---|---|
| batch | text | $6 | $6 | — | seller |
| batch | text · long | $1 | $1 | — | seller |
| batch | text · short | $1 | $1 | — | seller |
| standard | text | $12 | $12 | — | seller |
| standard | text · long | $2 | $2 | — | seller |
| standard | text · short | $2 | $2 | — | seller |
per million tokens, US dollars. * marks a rate from a single registry or two that disagree — the best available number rather than a confirmed one. Every other row is confirmed against a seller or a licensed index.
What it scores, by house
| Benchmark | House | Score | Evaluated | Scaffold |
|---|---|---|---|---|
| metr time horizons | collected by Epoch AI | 0.1 | — | — |
These scores are not comparable between houses and are not added up. A score is dated by when it was evaluated, not when the model was released, and a score without its scaffold is not a measurement — which is why both are printed.
What retires, and when
- deprecation — 2024-01-04, replaced by gpt-3.5-turbo-instruct
- deprecation — 2024-01-04, replaced by gpt-3.5-turbo-instruct
- deprecation — 2024-10-28, replaced by gpt-4o-mini
- deprecation — 2026-09-28, replaced by gpt-5.6-terra
- deprecation — 2026-10-23, replaced by gpt-5.6-terra
Every rate we hold · Why there is no single best model · Everything from OpenAI