Anthropic is an artificial intelligence security and research company located in San Francisco, USA. It operates as a Public-Benefit Corporation and is committed to building cutting-edge AI systems that are reliable, explainable, and controllable. The company's most well-known product is the Claude series of large language models, and it is at the forefront of the industry in AI alignment, interpretability and security research.

Creation and Team

2021 Co-founded by Dario Amodei and Daniela Amodei, brother and sister, and many other former OpenAI core researchers.
Propose alignment methods such as Constitutional AI, emphasizing honesty, harmlessness and self-correction.
MCP (Model Context Protocol, Large Model Context Protocol) was released in November 2024, becoming an open standard for connecting large models and external tools, ushering in the AI Agent era.

Flagship model

As of mid-2026, the Claude family is stratified by capabilities and costs:

Opus: The most powerful flagship, the latest is Claude Opus 4.8, which performs well in agent (Agentic) tasks and advanced programming
Sonnet: the balanced choice, latest Claude Sonnet 4.6
Haiku: high speed and lightweight, the latest is Claude Haiku 4.5
Mythos-class: Claude Fable 5 (June 2026) is the first Mythos-class model

Latest Development (2026)

Completed US$65 billion Series H financing, valued at approximately US$965 billion, surpassing OpenAI to become the world's most valuable AI startup.
The enterprise-level business is growing rapidly, with annual run-rate revenue exceeding approximately 47 billion US dollars.
June 1, 2026 has secretly submitted an IPO application and signed large-scale computing power agreements with Amazon, Google, Broadcom, etc.

Leader: Dario Amodei (CEO) Official website: anthropic.com

Note: The financing, valuation and revenue data in this section are public information as of mid-2026 and may change over time.