Model Overview
o3 is OpenAI's most powerful reasoning model, setting a new standard for math, science, coding, and visual reasoning tasks. It excels at technical writing and instruction-following, and is designed for multi-step problems that involve analysis across text, code, and images.
Key Features
- 200,000 context window
- 100,000 max output tokens
- June 1, 2024 knowledge cutoff
- Well-rounded and powerful across domains
- Excels at technical writing and instruction-following
- Designed for multi-step problems involving text, code, and images
Technical Specifications
- Input price: $10.00 per million tokens
- Output price: $40.00 per million tokens
- Supports: Input: text and image (implied), Output: text only
- Features: Function calling, structured outputs supported; Distillation and predicted outputs not supported
Snapshots
Positioning and Use Cases
Positioned as OpenAI's most powerful reasoning model, o3 is ideal for complex tasks requiring advanced reasoning capabilities. Best suited for multi-step problems that involve analysis across text, code, and images, as well as technical writing and instruction-following scenarios.
Official Documentation
