Anthropic has opened a public beta of self-hosted environments for Claude Code, letting organizations run coding-agent cloud sessions on servers they control instead of Anthropic's infrastructure. The option is available to Claude Team and Enterprise customers and is off by default.
Anthropic announced the move on August 6, 2026. A session started from the web, mobile, desktop, the terminal, or a scheduled routine runs inside the organization's network, alongside its internal services, toolchains, and security controls. Anthropic said it still steers most enterprises to its hosted offering, and positions self-hosting for teams whose network, tooling, or compliance requirements call for keeping agent execution on infrastructure they own.
How data is split
The data split is specific. Repository checkouts, build artifacts, secrets, and any files a session creates or changes stay on machines the organization provisions. The conversation itself, including prompts, responses, and tool results, is sent to Anthropic's API for model inference, and Anthropic stores the session transcript so a session can be resumed from any surface.
Organizations create named environments in the claude.ai admin settings, each grouping a set of runners: long-lived processes deployed on hosts inside the network that work like self-hosted CI runners. When a developer starts a cloud session and picks an environment, Anthropic's control plane queues it, a runner claims it, clones the repository from GitHub using credentials the deployment provides, and launches a Claude Code process to run the task.
Runner modes and network architecture
Runners run in two modes. Fixed mode keeps a set number running and spreads sessions across them; on-demand mode uses an orchestrator that starts runners as sessions arrive and stops them when work finishes. A runner serves one user at a time, locking to that user's account when it claims its first session so checked-out code does not mix between users. If a runner stops polling for roughly 60 seconds, the server requeues the session to another runner.
Every connection between the environment and Anthropic is outbound HTTPS to api.anthropic.com, covering queue polling, the session's event stream, and model inference. Anthropic never connects into the customer's network, and corporate egress proxies are supported.
Availability and limits
An owner or admin must enable the feature from the cloud-environments admin page, which requires Claude Code on the web turned on for the organization. Organizations using Anthropic's zero-data-retention setting cannot use it. Inference cannot be routed through Amazon Bedrock, Google Cloud's Agent Platform, or Microsoft's Foundry or an LLM gateway. Sessions started from Claude Tag, Claude Security, and Code Review do not route to self-hosted environments yet.
The launch extends a run of enterprise moves from Anthropic in the first week of August. Two days earlier the company said it is building an in-house silicon team for Claude.
Sources
Claude Code Sessions Can Now Run on Infrastructure Your Team Controls - Unite.AI: https://www.unite.ai/claude-code-sessions-can-now-run-on-infrastructure-your-team-controls/
Run Claude Code sessions on your own compute - Anthropic: https://claude.com/blog/run-claude-code-sessions-on-your-own-compute
Self-hosted environments - Claude Code docs: https://code.claude.com/docs/en/self-hosted-environments
