Model · OpenAI · Reasoning
o3
OpenAI's flagship reasoning model. Uses extended chain-of-thought before answering, trading latency for depth on math, science, and complex coding tasks.
- Modality
- Reasoning
- License
- Proprietary (Proprietary)
- Context window
- 200,000 tokens
- Released
- April 16, 2025
- Last verified
- May 10, 2026
- Runs locally
- No
Strengths
- Top-tier on math, scientific reasoning, and competition-level coding
- Strong tool use during reasoning chains
Weaknesses
- High latency — not suitable for chat-paced UX
- Closed weights
- Reasoning tokens are billed and can dominate cost
Try it
| Where | Type | Notes |
|---|---|---|
| OpenAI API | hosted-api | API key required |
| ChatGPT | product | Plus/Team/Enterprise |
Version history
- o3 Apr 2025 Current
- o4-mini Apr 2025
Official sources
- Model docs docs
Change log
- — Initial entry.
Esc