- Managed Agents runs Claude in an Anthropic-managed sandbox with
bash, file tools, and any custom tools you declare. - Browserbase provides the isolated Chromium session the agent connects to over CDP, plus Live View, session recordings, and Agent Identity.
Managed Agents is in beta. Requests require the
anthropic-beta: managed-agents-2026-04-01 header. See Anthropic’s docs for current status.Why pair them
Claude needs two things to browse the web reliably: a durable runtime to hold the loop, and a real browser to drive the page. Running either one yourself is painful.- Anthropic owns the runtime. No Lambda cold starts, no container orchestration, no token streaming plumbing. Dispatch a prompt and Anthropic runs the agent loop to completion.
- Browserbase owns the browser. No Chromium pinning, no fingerprint maintenance, no crashed-session cleanup. Every run gets a fresh, isolated sandbox with observability baked in.
- The split keeps state clean. Managed Agents holds the conversation and tool outputs. Browserbase holds the page, cookies, and downloads. You can inspect either side independently.
How it works
Create the agent and environment
You call
POST /v1/agents and POST /v1/environments once to register the system prompt, toolset, and npm packages Claude will have available. Anthropic returns an agent_id and environment_id you reuse across dispatches.Dispatch a run
For each user request, you call
POST /v1/agents/{id}/dispatch with the prompt and a reference to the environment. Anthropic spins up the sandbox, installs the declared packages, and starts the loop.Claude creates a Browserbase session
Inside the sandbox, Claude creates a session — either with
curl directly against the Browserbase REST API or with bb sessions create from the Browserbase CLI. Both return a connectUrl and a sessionId.Claude drives the browser
Claude issues
browse commands against the session — open, screenshot, snapshot, click, type — over CDP. You watch progress in Live View or stream events back to your own UI through a custom tool.What the stack looks like
| Layer | Owner | Role |
|---|---|---|
| Agent loop, tool execution, conversation state | Anthropic Managed Agents | Runs Claude to completion in a sandbox |
bash, file tools, custom tools | Anthropic Managed Agents | Built-in toolset, plus any custom tool you declare |
browse CLI or Stagehand | Your code (runs inside the sandbox) | Translates Claude’s intent into CDP actions |
| Chromium session, Live View, recordings | Browserbase | Isolated browser with observability and Agent Identity |
When to use this
Use Managed Agents with Browserbase when:- You want Claude to browse the web but don’t want to host the agent loop yourself.
- You need persistent conversation state across many tool calls (Slack bots, long research tasks, or queued jobs).
- You want Anthropic-side tool-use traces for debugging alongside Browserbase session recordings.
- You’re building a chat surface (Slack, Discord, or a web app) where each thread is a long-running agent run with steering.
- You’re already building a general-purpose Managed Agent and want to add web browsing as another capability alongside its existing tools.
Getting started
Quickstart
Create a Managed Agent, dispatch a run, and drive a Browserbase session end to end.
Browse CLI
The lightweight entry point Claude uses inside the sandbox to control Browserbase.
Further reading
Managed vs self-hosted
Why managed Browserbase sessions beat self-hosted Chromium in production.
Session Live View
Watch a Managed Agent drive a browser in real time.
Agent Identity
Verified browsers, proxies, and authentication for agents on the open web.