88Z AI, 13 November — Anthropic has announced a private beta for self-hosted Claude Managed Agents, allowing enterprises to deploy the agent harness, sandboxes, and session history in their own cloud or on-premises environments. This solution privatizes the orchestration layer rather than running the model locally, with model requests still directed to Anthropic or customer-chosen providers, addressing security needs for accessing internal code and systems.
At a glance
- Supports self-hosting of agent orchestration, sandboxes, and history, not the model itself
- Ideal for tasks needing access to internal code and systems, such as build fixes
- Eligible organizations can request private beta access via their Anthropic account team

Core Features and Deployment Model
Anthropic’s official developer account announced that Claude Managed Agents are now in private beta. Enterprises can host the agent harness, sandboxes, and session history in their own cloud or on-premises infrastructure, choosing where to send Claude model requests. Crucially, this setup privatizes the orchestration layer only, not the underlying Claude model itself.
Use Cases and Access
Designed for tasks requiring access to internal codebases and systems, such as fixing broken builds or triaging alerts, this feature is currently available only to eligible organizations via request through their Anthropic account team. Model requests can still be sent to Anthropic or routed to customer-chosen providers like Amazon Bedrock, Google Agent Platform, or private gateways.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- @ClaudeDevsSource published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.