o3-mini
Reasoning model Generally availableOpenAI high-efficiency reasoning model offering configurable chain-of-thought deliberation for fast, cost-effective STEM, coding, and logical tasks.
Overview
OpenAI o3-mini is a highly efficient reasoning model released in January 2025 as part of OpenAI's next-generation o3 series. Engineered for STEM research, software engineering, and analytical problem-solving, o3-mini delivers near-flagship reasoning performance with significantly reduced latency and operating costs.
Supporting a 200,000-token context window and up to 100,000 completion tokens, o3-mini introduces configurable reasoning effort parameters ('low', 'medium', 'high') allowing developers to balance latency against reasoning depth depending on query complexity. It provides native support for developer tools including function calling, structured output generation, developer messages, and text reasoning.
o3-mini is accessible via OpenAI's Direct API and partner API platforms. It operates as a proprietary hosted service with no open weights available for self-hosting.
Technical Specifications
Context Window
200,000 tokens
Max Input Tokens
200,000
Max Output Tokens
100,000
Architecture
Reasoning Transformer
Released
2025-01-31
Modalities
Text → Text (Both)
Capabilities
Text Generation
Code Generation
Reasoning
Function Calling
Long Context
Use Cases
Coding Assistance
Data Analysis
Research
Languages
English (en)
Spanish (es)
French (fr)
German (de)
Chinese (zh)
Access Routes
| Route | Type | Protocol | Region / Scope | Provider | Status |
|---|---|---|---|---|---|
| OpenAI API | Hosted API (Direct provider) | OpenAI-compatible REST API | Global | OpenAI | Available |
Provider
Identifiers
- o3-mini Canonical identifier Default
- o3-mini-2025-01-31 Canonical identifier Default