Anthropic
frontierUS · Founded 2021
Safety-focused frontier lab. Makes Claude.
Anthropic was founded by former OpenAI employees with a safety-first mandate. Constitutional AI, the training technique that makes Claude's refusals specific rather than vague, was Anthropic's primary research contribution. The lab operates on the thesis that making models safe and making them capable are the same problem.
Main models
- Claude Opus 4.7
- Claude Sonnet 4.6
- Claude Haiku 4.5
Strengths
- Safety research
- Long-context models
- Agentic discipline
- Constitutional AI
Pricing
$1–$75 per million tokens depending on model and tier