Definition
Anthropic is an American AI company that develops the Claude family of large language models. It was founded in 2021 by former OpenAI researchers, including CEO Dario Amodei and President Daniela Amodei, and is organized as a public benefit corporation with AI safety research at the center of its stated mission.
As of October 2026, Anthropic sells several model lines. Claude Fable 5.1, released on September 1, 2026, is its most capable model open to all customers, while Claude Mythos 5.1 is limited to participants in its Project Glasswing program. The Claude 5.5 generation arrived over three weeks: Opus 5.5 on September 22 for long-running agentic coding and knowledge work, Sonnet 5.5 on September 28 as the everyday balanced model, and Haiku 5.5 on October 7 for high-volume, low-latency tasks. Opus 5.5 and Haiku 5.5 both ship with a 1 million token context window. The models are available in the Claude apps, through the Claude API, and on Amazon Bedrock, Google Cloud and Microsoft Foundry, and they power Claude Code, Anthropic's coding agent.
The company has grown quickly. Anthropic raised $30 billion at a $380 billion valuation in February 2026 and $65 billion at a $965 billion post-money valuation in May 2026, when it said run-rate revenue had passed $47 billion. On June 1, 2026 it confidentially submitted a draft S-1 to the SEC for a proposed IPO.
Two pieces of Anthropic's work shaped the wider industry. The first is constitutional AI, a training method in which a model critiques and revises its own outputs against a written set of principles, reducing reliance on human labels compared with pure RLHF. It is part of Anthropic's broader research on AI alignment and interpretability. The second is the Model Context Protocol (MCP), an open standard for connecting AI applications to tools and data that Anthropic released in November 2024. By December 2025, when Anthropic donated MCP to the Linux Foundation's Agentic AI Foundation, it counted more than 10,000 active public MCP servers and adoption in ChatGPT, Cursor, Gemini, Microsoft Copilot and Visual Studio Code.
For GEO, Anthropic matters in two ways. Claude answers questions with web search, so brands can be cited in its responses the same way they are in ChatGPT or Perplexity. And Anthropic runs three distinct web robots, documented in its crawler help article: ClaudeBot collects content that may be used for training, Claude-SearchBot indexes pages to improve search results, and Claude-User fetches pages when a person asks Claude a question. Blocking ClaudeBot opts future content out of training but leaves search and user fetches untouched, so each should be decided separately (see citation crawlers vs training crawlers). Promptwatch's crawler data shows Claude's citation crawler grew more than 100x as a share of tracked AI citation-crawler traffic between December 2025 and June 2026, so Claude-User requests are now a visible share of AI bot traffic on many sites.
Examples of Anthropic
- A documentation team sees Claude-User requests climb in its CDN logs after Claude Code users start asking the agent to read its API reference, and it adds a plain-text changelog page so the agent finds current versions.
- A news publisher blocks ClaudeBot to keep its archive out of training data but leaves Claude-SearchBot and Claude-User allowed, so its articles can still be cited when Claude searches the web.
- A fintech company picks Claude Haiku 5.5 for a customer-support classifier because of its price and latency, while its engineering team uses Opus 5.5 inside Claude Code for refactoring work.
- A SaaS vendor publishes an MCP server for its product so that users of Claude, ChatGPT and Cursor can query their account data from inside those assistants.
