Anthropic
Claude 4.5, Claude 4, Claude 3.5 Sonnet, and Haiku
✓ What's Allowed
- ✓Building applications using your own API key
- ✓Commercial products and SaaS platforms
- ✓Research, education, and internal tools
- ✓Content generation with responsible use
- ✓Integrating Claude into your product workflow
✗ What's Prohibited
- ✗Reselling the Services without express approval
- ✗Generating harmful, deceptive, or illegal content
- ✗Developing competing AI models without express approval
- ✗Attempting to extract model weights or architecture details
- ✗Circumventing safety features or usage limits
- ✗High-Risk consumer-facing use (legal, healthcare, etc.) without the required qualified-professional review and AI disclosure
Key ToS Sections
Usage Policy (also called the AUP)View →
Universal Usage Standards, High-Risk Use Case Requirements and Additional Use Case Guidelines
Commercial Terms A.1, D.4View →
Customer applications are permitted; resale requires express approval
Commercial Terms B (Customer Content)View →
Customer retains rights to Inputs and owns Outputs; Anthropic may not train models on Customer Content from the Services
Legacy editorial date (not policy verification): 2025-10-16 —Check your compliance →