Key Features
• Advanced Reasoning Capabilities
-1. Optimized for STEM tasks including science, math, and coding
-2. Matches or exceeds o1 performance in reasoning benchmarks
-3. Supports three reasoning effort levels: low, medium, and high
• Cost Efficiency
-1. 90% cheaper than o1, priced at $0.55 per million input tokens and $4.40 per million output tokens
-2. Competitive pricing compared to models like DeepSeek R1
• High Performance
-1. Outperforms o1 in coding tasks and select reasoning benchmarks
-2. 24% faster response time than o1-mini, averaging 7.7 seconds
• Developer-Friendly Features
-1. Supports function calling, structured outputs, and developer messages
-2. Compatible with Chat Completions API, Assistants API, and Batch API
-3. Streaming and tool integration for real-world applications
• Search Integration
-1. Early prototype for web search, providing up-to-date answers with source links
• Accessibility
-1. Available to all ChatGPT users: free tier with limits, Plus/Team at 150 messages/day, Pro with unlimited access
-2. Integrated into Microsoft Azure OpenAI Service, GitHub Copilot, and GitHub Models
