Deploy OpenClaw on Local Machine

Run OpenClaw on Local Machine with AgentSea — the easiest way to deploy Grid-backed AI agents on infrastructure you control. Copy the command below: it installs the CLI, then walks you through deployment interactively.

AgentOpenClaw

EnvironmentLocal Machine

Run this command

Paste this in your terminal (macOS, Linux, or WSL). It installs AgentSea, then starts agentsea openclaw local and guides you through the rest — API keys, provisioning, and connecting to your agent.

shell

curl -fsSL https://agentsea.thegrid.ai/install.sh | bash -s -- openclaw local

Already installed? Run deploy only

shell

agentsea openclaw local

Install CLI without deploying

shell

curl -fsSL https://agentsea.thegrid.ai/install.sh | bash

Install options and flags in the CLI Reference.

Common commands

shell

agentsea                              # Interactive picker
agentsea ls                            # List your deployments
agentsea matrix                        # Agent x cloud matrix

After install — what to do next

Once AgentSea finishes provisioning OpenClaw on Local Machine, here's how to get the most from your agent.

Open the Control UI — AgentSea starts the gateway and opens it in your browser (SSH tunnel on cloud servers).
Chat in the terminal TUI (openclaw tui) or use the web Control UI — both talk to the same agent.
Connect messaging channels like Telegram, WhatsApp, or Discord through the gateway. Channels overview
Your heartbeat model keeps the agent responsive between messages — already configured by AgentSea.
Reconnect anytime with agentsea connect or agentsea list. OpenClaw docs

Prefer not to install the CLI?

Run this one-liner instead — no global install required. It downloads and runs the bootstrap script for OpenClaw on Local Machine.

shell

bash <(curl -fsSL https://agentsea.thegrid.ai/local/openclaw.sh)

← Pick another agent | Change environment