o3
Reasoning model that spends additional compute at inference time before answering.
Source: OpenAI announcement · Updated 16 Apr 2025 · provider announcement
Capabilities
Capability ratings
Editorial assessment of what the system can do on each axis, 0-100. Capability is not adoption and not employment impact.
- Code88
Summary
Extended chain-of-thought reasoning for maths, science and code; tool use during reasoning.
Strongest axis for this model.
Average across all rated axes.
Published benchmarks
| Benchmark | Score |
|---|---|
| AIME 2024 | 96.7 (with tools) |
Only provider-published figures are stored. This site does not run its own benchmarks.
Impact
Occupations affected by o3 (0)
Occupations with published evidence mapped to this model. A model being able to do something is not evidence that firms use it.
No occupation-level evidence has been mapped to this model yet. This is a capability record only.
Provenance
Where this record came from
Release records cite the provider announcement. Open-weight releases also link to their public model and repository pages where available. Read /methodology for how capability ratings are produced.