Choose your agent
Start with a focused template or bring your own OpenClaw setup.
OpenClaw is free and you can install it yourself. What you cannot install is a spend cap that refuses, a switch that stops the whole fleet, and a log that records what your agent was not allowed to do.


Claw is not here to give you another plan. It connects to your stack, does the work, and returns something you can use.
Dashboards, briefs, follow-ups, and shipped work, not another answer to copy into a new tab.
Connect the channels and tools you already use. Claw carries context between them.
Schedule checks, catch changes, and return with the result before the next meeting.
Persistent memory keeps your preferences, formats, and decisions in the loop.
Claw turns a plain-English request into real work across your tools. See what it is doing, approve what matters, and get the finished result back in your channel.
See the product ↗Your agents handled the work overnight.

Claw handles the infrastructure so you can focus on what your agent actually does. Start small. Scale when you're ready.
Start with a focused template or bring your own OpenClaw setup.
Give it access to the channels and tools your work already lives in.
Your agent runs persistently, picks up context, and delivers the outcome.
From personal automation to a whole AI workforce, your agents are always on and ready to move work forward.

Inbox zero, daily briefings, reminders, research, and the small things that add up.
For individuals →
Support, sales ops, content, and engineering agents that plug into your team.
For teams →
Governed, observable AI operations with the controls your business requires.
For enterprise →A spend cap that refuses the run, a kill switch that stops the whole workspace, and an approval gate that holds work until a person decides. All three are checked on the server before OpenClaw is called.
See what we enforce, and what we do notEvery plan gives you a persistent OpenClaw runtime, credits you can count yourself, and a spend cap on every agent. Claw Install is a one-time $2,500 if you want the fleet standing correctly on day one, and Enterprise runs on an annual contract.
For one person running an agent as part of real work. 3 agents, 500 run credits, spend cap per agent, and the full audit log.
Start with Operator ↗For ops and engineering teams whose agents touch production. 15 agents, 5,000 credits, role controls, exportable audit, egress allowlist.
Start a Team workspace ↗For agencies and MSPs running agents for many clients. 60 agents, 20,000 credits, isolated client workspaces, per-client reporting.
Start with Partner ↗Operator is $49 a month. Every agent starts with a spend cap you control.