Claude AI Explained: Anthropic's Assistant and How It Works
Claude is Anthropic's family of large language models. It is the company's primary commercial product and the basis of most of its reported revenue.
- Published
- Updated
- Author
- Anthropic Guide Research Desk
01What Claude is
Claude is a family of large language models developed by Anthropic and first made publicly available in March 2023. The models generate text, analyse documents, write and review software, and are accessed through a consumer chat interface, a developer API and major cloud marketplaces.
Anthropic positions Claude as a general-purpose assistant with a deliberate emphasis on reliability and refusal behaviour in sensitive contexts, rather than on maximal capability alone.
02How Anthropic describes its training approach
Anthropic has published work on what it calls Constitutional AI, an approach in which model behaviour is shaped against an explicit written set of principles rather than relying solely on human preference labelling. The company has published the principles it uses and updates them over time.
This is a company-stated methodology. Independent evaluation of model behaviour is carried out by third-party researchers and by customers, and results vary by task.
03Where Claude is used
Claude is distributed directly by Anthropic and through cloud partners, including Amazon Bedrock and Google Cloud. Reported enterprise use concentrates on software engineering assistance, customer operations, document analysis and internal knowledge tools.
