OpenAIdavinci-002
davinci-002 is an OpenAI proprietary model, released 15 Mar 2022. It is not yet ranked: no independent benchmark results so far, only pricing and metadata.
- Blended price
- –
- Output speed
- –
- First answer
- –
- Context
- –
- Released
- 15 Mar 2022
Versions
OpenAI has shipped 4 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| davinci-003 | 28 Nov 2022 | – | – |
| davinci-002this page | 15 Mar 2022 | – | – |
| davinci-002 | 15 Mar 2022 | – | – |
| Davinci 002 | 28 May 2020 | – | – |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | not stated | Source |
|---|---|---|
| HellaSwagnot in index | 81.5%#24 | Epoch AI Benchmarking Hub |
| WinoGrandenot in index | 81.6%#12 | Epoch AI Benchmarking Hub |
| ARC (AI2 Challenge)not in index | 85.2%#12 | Epoch AI Benchmarking Hub |
Maths
| Benchmark | not stated | Source |
|---|---|---|
| GSM8Knot in index | 41.5%#44 | Epoch AI Benchmarking Hub |
Knowledge
| Benchmark | not stated | Source |
|---|---|---|
| MMLUnot in index | 70.0%#53 | Epoch AI Benchmarking Hub |
See also
Data as of 19 Sept 2026. Compare with another model.
Cite as: BenchLeader, “davinci-002: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/text-davinci-002, data as of 19 Sept 2026.