Anthropic's new Claude Enterprise Plan aims to help companies securely collaborate with its AI using internal knowledge. The plan features an expanded context window, increased usage capacity, and native GitHub integration.
The Enterprise Plan offers a 500,000 token context window, higher usage limits, and native GitHub integration for working with Claude on entire codebases. Anthropic claims users can provide Claude with knowledge equivalent to hundreds of sales logs, dozens of 100+ page documents, or medium-sized codebases.
However, companies shouldn't rely too heavily on precise processing of large documents, as large language models tend to ignore information in the middle of documents or overemphasize information at the beginning - a phenomenon known as "lost in the middle." This issue also affects other AI models with massive context windows, like Gemini 1.5.
The Enterprise Plan includes security features such as single sign-on (SSO), role-based permissions, and admin tools to protect data and teams. Anthropic emphasizes that customer data is protected: Claude is not trained on users' conversations or content. Audit logs and SCIM are expected to be available in the coming weeks.
AI for code: Anthropic targets Cursor
The native GitHub integration allows development teams to sync GitHub repositories with Claude to collaborate on new features, fix bugs, or onboard new developers.
GitHub integration is currently available in beta to early Enterprise Plan adopters and will be more widely released later this year. It's the first of several native integrations Anthropic plans to build to connect Claude with enterprise data sources.
With this GitHub integration, Anthropic is competing with AI code interface Cursor.ai, which has gained many customers thanks to the coding capabilities of Anthropic's Sonnet 3.5 AI model. Anthropic likely wants a piece of this market.
Companies interested in the Enterprise Plan should contact Anthropic's sales team.
