claudekit / updates / run-claude-code-sessions-on-your-own-compute
[ NEW · ]

Run Claude Code sessions on your own compute with self-hosted environments

Now in public beta, self-hosted environments let you run Claude Code sessions on your own infrastructure. You operate runners in either fixed or on-demand mode, and each session runs in its own checkout so work stays isolated between developers and accounts. Repository checkouts, build artifacts, and secrets stay on infrastructure you provision, while the conversation is sent to Anthropic for inference. Available to organizations on Claude Team and Enterprise plans.

Official announcement →

This article is a summary based on official documentation.

Overview

Now in public beta, self-hosted environments let you run Claude Code sessions on your own infrastructure. Sessions from every supported surface route to the same environment, so you set it up once and it works wherever your team starts a session.

Key features

  • Runners you operate

    Claude Code sessions have run on Anthropic-hosted compute. With self-hosted environments, your organization runs the runners that execute sessions. Self-hosted environments are for teams whose network, tooling, or compliance requirements call for keeping agent execution on infrastructure they control.

  • Internal network access and pre-installed tooling

    Sessions run inside your network and can reach internal services, databases, and registries without exposing them to the public internet. You can pre-install compilers, SDKs, and internal CLIs in your environment so every session starts ready to build.

  • Fixed and on-demand runner modes

    Runners come in two modes. Fixed: you keep a set number running and sessions are distributed across them. On-demand: an orchestrator watches for queued sessions, starts a runner as sessions arrive, and stops them when work finishes so capacity tracks demand.

  • Per-session checkouts

    Runners can serve more than one session, but each session runs in its own checkout, so work stays isolated between developers and accounts.

  • Where data lives

    Repository checkouts, build artifacts, secrets, and any files a session creates or modifies all stay on infrastructure you provision. The conversation itself, including prompts, responses, and tool results (which can include code that Claude reads), is sent to Anthropic for inference, and the session transcript is stored so a session can be picked up from any surface.

  • One environment for every surface

    Sessions from every supported surface route to the same environment, so you set it up once and it works wherever your team starts a session.

Notes

  • Availability — self-hosted environments are available in public beta to organizations on Claude Team and Enterprise plans. They are off by default and not available for organizations using ZDR.
  • Someone has to own it — plan on a platform, developer experience, or developer productivity team owning setup and ongoing operation, including building and maintaining the runner image, updating runners, and running the orchestrator if you use on-demand mode.
  • Not the same as Remote Control — sessions using Remote Control end when that machine stops running the session and are tied to the user who ran claude, whereas self-hosted environments run sessions on shared infrastructure your platform team operates and can be used by any user.
  • Setup details — see the documentation to learn more.
§ 4

Frequently Asked Questions

frequently asked
§ 4.1
What are self-hosted environments?
Now in public beta, self-hosted environments let you run Claude Code sessions on your own infrastructure instead of Anthropic-hosted compute.
§ 4.2
Which plans can use them?
Self-hosted environments are available in public beta to organizations on Claude Team and Enterprise plans. They are off by default and not available for organizations using ZDR.
§ 4.3
Does my code get sent to Anthropic?
Repository checkouts, build artifacts, secrets, and any files a session creates or modifies all stay on infrastructure you provision. The conversation itself, including prompts, responses, and tool results (which can include code that Claude reads), is sent to Anthropic for inference, and the session transcript is stored so a session can be picked up from any surface.
§ 4.4
How do runners work?
Runners come in two modes. In fixed mode you keep a set number running and sessions are distributed across them. In on-demand mode an orchestrator watches for queued sessions, starts a runner as sessions arrive, and stops them when work finishes so capacity tracks demand.
§ 4.5
Can one runner serve multiple developers?
Runners can serve more than one session, but each session runs in its own checkout, so work stays isolated between developers and accounts.
§ 4.6
How is this different from Remote Control?
Sessions using Remote Control end when that machine stops running the session and are tied to the user who ran `claude`, whereas self-hosted environments run sessions on shared infrastructure your platform team operates and can be used by any user.