Anthropic Opens Self-Hosted Claude Managed Agents in Private Beta

Anthropic announced on November 12 that Claude Managed Agents is now available as a self-hosted edition in private beta. Enterprises can run the entire agent service in their own cloud or on-premises infrastructure, including the server that runs the agent loop, the sandboxes where agents execute code, the credentials agents use, and the full history of every agent session.

What moves home, and what does not

A self-hosted deployment has three customer-operated pieces: the cma-server, which exposes the same Managed Agents API as the hosted service, runs the agent loop and stores agents, sessions and their histories; a PostgreSQL database, version 16 or later, that holds that history — without it, the server keeps data only in memory and a restart wipes it; and the workers that run agents' shell and file tools inside the customer's sandboxes.

The one thing that does not move home is the model itself. Every request to Claude still goes to an inference provider the customer picks — Anthropic by default, or Amazon Bedrock, Google's agent platform, Microsoft Foundry, or the customer's own model gateway. Anthropic says it has no remote access to a customer's self-hosted server or its logs.

Background and limits

Claude Managed Agents is Anthropic's service for hosting autonomous Claude agents that work through long jobs on their own: fixing broken builds, testing releases, triaging alerts and working through bug backlogs. It launched as a hosted service on April 9, 2026, and self-hosted sandboxes were added in May, moving an agent's tool execution onto customer hardware while Anthropic kept running the agent itself. The November 12 announcement brings the rest in-house too.

The self-hosted edition is not yet a feature-for-feature copy of the hosted service: scheduled deployments, dreams and webhooks are unavailable, and web search and web fetch stay switched off until a team explicitly enables them. Anthropic describes a Docker Compose setup for development and warns that it is not appropriate for production, pointing teams toward Helm-based deployment tooling. The private beta is open only to eligible organizations through Anthropic's account team; the company has published no pricing, no list of covered countries and no date for general availability.

For enterprises in regulated industries, the change removes a long-standing compliance obstacle: agent credentials and session history can stay inside their own security perimeter. The trade-off is operational — each release must be installed and maintained by the customer.

(File photo: Anthropic office exterior. Image: Despertar Matinal)