OpenClaw is a self-hosted Gateway that connects chat applications to AI coding agents. It keeps sessions, routing, and channel connections in one runtime, so a message can trigger an agent, approved tools can perform the work, and the result can return to the same conversation.
Antihost offers VPS plans with one-click OpenClaw, NVMe storage and a 14-day free trial.
The dependable setup is: install a supported Node.js release, run onboarding, install the Gateway as a service, verify it on port 18789, open the Control UI, send a test message, and then add a channel such as Telegram. From there, combine schedules, webhooks, hooks, skills, plugins, browser tools, device nodes, and multi-agent routing into controlled workflows.
As an Amazon Associate I earn from qualifying purchases.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesWhat OpenClaw does
OpenClaw’s official documentation describes it as a “self-hosted gateway that connects your favorite chat apps — Discord, Google Chat, iMessage, Matrix, Microsoft Teams, Signal, Slack, Telegram, WhatsApp, Zalo, and more via channel plugins — to AI coding agents.” The Gateway is the single source of truth for sessions, routing, and channel connections. You run it on a computer or server you control; channel adapters receive messages, the selected agent performs work, and the response is delivered through the originating channel.
After startup, the local Control UI is available at http://127.0.0.1:18789/. A foreground process is useful during first-time setup, but an installed service is the practical choice for an always-on assistant.
#1 Best Overall
- Includes Raspberry Pi 5 with 2.4Ghz 64-bit quad-core CPU (8GB RAM)
- Includes 128GB Micro SD Card pre-loaded with 64-bit Raspberry Pi OS, USB MicroSD Card Reader
- CanaKit Turbine Black Case for the Raspberry Pi 5
- CanaKit Low Noise Bearing System Fan
- Mega Heat Sink - Black Anodized
Check prerequisites before installing
Use a supported Node.js version
Install Node.js 24.16 or newer, or 26.1 or newer. Node 26 is the recommended line. Check the version that will actually run in your shell:
node --versionIf the command reports an older release, upgrade Node before onboarding. Re-check compatibility after OpenClaw or Node updates because supported versions and defaults can change.
Recommended Free Tools
Prepare an AI provider login
Onboarding can reuse an existing Claude Code or Codex CLI login, or you can provide an API key from a supported provider. Have the credential available before you start. A detected login is not proof that every model route works; the Quick start flow verifies access with a real completion.
Choose where the Gateway will live
A desktop is simplest for an interactive trial. A Linux service, Windows Hub, WSL2 environment, VPS, or a small always-on computer is better when automations must continue while your main workstation is asleep. Decide whether the Gateway will be reachable only from the host or from other networks before connecting channels or enabling host tools.
Install OpenClaw
macOS or Linux installer
Run the documented installer from a terminal:
curl -fsSL https://openclaw.ai/install.sh | bashReview what the script is doing in your environment and ensure the resulting openclaw command is on your PATH. Open a new shell if your shell profile was modified.
Windows PowerShell installer
In PowerShell, use:
iwr -useb https://openclaw.ai/install.ps1 | iexRun the command in the account that will own the Gateway service. Installing under one account and onboarding under another can produce confusing service and credential paths.
Other installation paths
The project also documents npx openclaw@latest, npm, pnpm, bun, Docker, Nix, desktop companion applications, WSL2, and installation from source. For a source checkout, the documented development sequence is:
Rank #2
- Includes Raspberry Pi 5 16GB with 2.4Ghz 64-bit quad-core CPU (16GB RAM)
- Includes 128GB Micro SD Card pre-loaded with 64-bit Raspberry Pi OS, USB MicroSD Card Reader
- CanaKit Turbine Black Case for the Raspberry Pi 5
- CanaKit Low Noise Bearing System Fan
- Mega Heat Sink - Black Anodized
git clone <repository>
pnpm install
pnpm build
pnpm ui:buildThe repository states that running plain npm install at its root is not supported. Use the package-manager path documented for the checkout you are using rather than mixing commands.
Run onboarding and prove provider access
Quick start
Choose Quick start when you want OpenClaw to reuse detected AI access. The wizard checks the route with a real completion, which catches invalid or expired credentials earlier than simply saving a token.
Custom setup
Choose Custom setup when you need to select a provider, model route, workspace, or other options explicitly. If the detected route fails, switch to manual provider setup instead of repeatedly retrying an unusable login.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Use the CLI wizards later
For the classic guided flow, run:
openclaw onboard --classicFor targeted changes after onboarding, run:
openclaw configureThe getting-started documentation describes initial setup as taking about five minutes. That is a documentation estimate, not a measured guarantee; credential approval, package downloads, firewall rules, and channel setup can take longer.
Install the Gateway as an always-on service
When onboarding starts the Gateway in the foreground, leave it running until you have confirmed the Control UI and a test completion. Then stop that process with Ctrl+C and install the service:
- Stop the foreground process. Press
Ctrl+Cin the terminal where onboarding left the Gateway running. - Install the service.
openclaw gateway install - Check its state.
openclaw gateway status - Open the dashboard.
openclaw dashboard - Send a test message. Use the Control UI and confirm that an agent responds before adding an external channel.
The service maps to a macOS LaunchAgent, a Linux or WSL2 systemd user unit, or a native Windows Scheduled Task. If Windows denies task creation, the documented fallback is a per-user Startup-folder entry. The status output should show the Gateway listening on port 18789. If the dashboard does not open, verify the service status and inspect logs before changing network settings.
Add a chat channel
Telegram is a practical first channel
The documentation identifies Telegram as the fastest first channel because it requires a bot token. Add it through the guided or scripted channel command:
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchRank #3
- CanaKit Raspberry Pi 5 Essentials Starter Kit
openclaw channels addFollow the prompts for the Telegram plugin and token, then send a message to the bot. Keep the first test in a private chat so you can verify pairing and permissions without involving other users.
Other supported channel families
The official channel list includes Discord, Slack, WhatsApp, Signal, iMessage, Google Chat, Microsoft Teams, Matrix, Mattermost, and additional plugin channels. The exact credential fields, media behavior, group controls, and setup screens depend on the channel plugin and can change. Use openclaw channels add for the channel you have selected and consult that plugin’s current instructions.
| Channel example | Credential noted in the setup material | Pairing and group considerations |
|---|---|---|
| Telegram | Bot token | Pair and test in a private chat first; other permission details depend on the plugin configuration. |
| Discord | Not stated | Use the plugin’s current credential and server-permission flow. |
| Slack | Not stated | Use the plugin’s current app and workspace flow. |
| Not stated | Use the plugin’s current account-linking flow. | |
| Signal, iMessage, Google Chat, Microsoft Teams, Matrix, Mattermost | Not stated | Capabilities and permission controls vary by plugin; verify them before enabling group access. |
Integration coverage is broad but volatile. The official integrations directory currently lists 29 chat channels, 64 model and media providers, and 142 official plugins (OpenClaw, 2026). These are page counts, not a permanent compatibility promise.
Assemble an end-to-end automation
OpenClaw does not impose one universal workflow. A useful design separates five decisions: what starts the run, which agent and session receive it, which tools are allowed, how the work is performed, and where the result is sent.
Free tools Windows power users keep installed
One-click scans. No signup required.
1. Select a trigger
- Chat message: A user sends a command in Telegram, Discord, or another connected channel.
- Schedule: A cron job starts a recurring task, such as a report or maintenance check. The CLI groups this area under cron and also uses the alias “automations.”
- Webhook: An external system calls an HTTP endpoint when an event occurs.
- Hook: An internal lifecycle event launches a follow-up action.
Choose one trigger per job and document its expected payload, schedule, and failure behavior. A webhook should not be treated as trusted merely because it comes from a familiar service.
2. Route to the right agent and session
Multi-agent routing can separate work by agent, workspace, or sender. Keep unrelated contexts apart: a code-maintenance agent should not automatically inherit a personal chat session, and a public group should not share a session containing private project data. Define the routing rule before granting tools.
3. Allow only the tools required
Skills provide repeatable procedures loaded on demand. Plugins add channel, provider, and tool surfaces. Browser tooling gives the agent a dedicated browser, while nodes expose device capabilities such as camera, screen, voice, location, and Canvas. Grant the smallest useful set: a scheduled status report may need a read-only API and a message destination, whereas a browser-based workflow may require a dedicated browser profile but not host shell access.
Rank #4
- All-in-One Complete Kit: This SANOOV RPi 5 bundle comes with Raspberry Pi 5 4GB RAM single board, active cooler, durable ABS case and screwdriver. No extra parts needed, ready to use right out of the box for beginners and hobbyists
- Powerful Single Board Computer: Equipped with 4GB RAM and high-performance processor, delivers fast running speed for 4K playback, AI projects, programming and daily computing tasks. SANOOV for raspberry pi 5 4GB is equipped with broadcom 64 quad-core Arm Cortex A76 processor with gigabit ethernet and upgraded with IEEE 802.11ac Wi-Fi, Bluetooth 5.0 dual-band 2.4Ghz and 5Ghz and Power Over Ethernet (POE). Upgrading delivers 2-3 x speed vs Pi 4, redefining the experience
- Efficient Active Cooler: Effectively lowers operating temperature and prevents performance throttling. Runs quietly even under long-time heavy load, ensures stable operation all day long. SANOOV RPi 5 4GB kit offer an active cooler, which combines an aluminium heatsink with a high-performance PWM fan. Active cooler is fully compatible with the Pi OS, which can effectively reduce the temperature of RPi5 and ensure its good performance during long-term high load operation
- Sturdy ABS Protective Case: Well-fitted for Raspberry Pi 5 board, can be secured with 4 screws to effectively protect the Pi 5 motherboard from damage, reserves full access to all ports and buttons. SANOOV uses ABS material to produce the case, which has a softer texture and feel. Meanwhile, SANOOV case adopts a layered design for easy disassembly and installation. (Tip: The Case cannot install M.2 HAT Add on Board and Solid State Drive!)
- Wide Application & Full Compatibility: Seamlessly compatible with official OS and mainstream peripheral accessories for Raspberry Pi 5. Whether you are a beginner, student, electronics hobbyist or professional developer, this all-in-one kit meets your diverse needs. It excels in IoT projects, robotics design, retro gaming devices, home media servers and other DIY creations. Backed by a large global community, you can easily find guides, technical support and shared projects online
4. Perform the work through a bounded procedure
Put stable instructions in a skill instead of relying on an improvised prompt each time. Have the procedure state inputs, allowed actions, success criteria, timeout behavior, and what to do when data is missing. Use a dedicated browser or node where possible so the automation does not silently operate in your personal browser session or expose unrelated device data.
5. Return the result and preserve a transcript
Send the outcome back to the originating channel or a designated operations channel. Include whether the run succeeded, failed, or needs approval. Keep transcripts and logs long enough to investigate errors, but apply your own retention and privacy policy to messages, screenshots, locations, and credentials.
Example pattern
A scheduled trigger can route to a reporting agent, allow a read-only skill and a browser tool, collect the requested data, and post a concise result to a private Telegram chat. A webhook can use the same pattern, but should authenticate the caller, validate the payload, prevent duplicate processing, and require approval before any destructive action. These are architecture patterns; OpenClaw’s documentation does not guarantee successful execution for a particular third-party service.
Choose a host for continuous operation
| Host | Setup effort | Always-on reliability | Network exposure | Cost and maintenance | Isolation and headroom |
|---|---|---|---|---|---|
| Local desktop | Lowest for first setup | Depends on sleep, reboots, and user availability | Can remain local; remote access is optional | Uses existing hardware; ordinary desktop maintenance | Usually the most memory and storage headroom; sandbox deliberately |
| Windows Hub or native Windows service | Moderate; service uses Scheduled Task or Startup fallback | Good while the machine is powered and signed in as configured | Local by default; expose only through an intentional access path | Uses existing PC; maintain Windows and the service | Host tools run on the machine unless sandboxing is configured |
| Linux service or WSL2 | Moderate; systemd user unit or WSL2 setup | Good on a dedicated, powered host | Can be private or remotely reachable; firewall it either way | Low hardware cost if reusing a machine; ongoing package updates | Good control over users, storage, and sandboxing |
| Raspberry Pi 5 | Moderate hardware and OS setup | Designed for low-power, continuous operation | Keep it behind a firewall or private network unless remote access is secured | Requires board, storage, power, and maintenance; no official performance benchmark is established here | Less memory and storage headroom than a desktop or VPS; keep workloads small and isolate tools |
| VPS or cloud host | Moderate to high, depending on provider and security model | Usually strongest for unattended uptime | Internet exposure is a primary concern; use firewalling, private access, and authentication | Recurring hosting cost plus server maintenance | Flexible storage and memory; easiest to dedicate a user, volume, and sandbox |
The integrations directory names Fly.io, Hetzner, GCP, Azure, exe.dev, and EasyRunner as VPS or cloud options. Their pricing, availability, and commercial terms are not established by the setup documentation, so select a host based on your region, budget, support needs, and security controls rather than assuming equivalence.
When a Raspberry Pi makes sense
A Raspberry Pi 5 is an optional small-host path for readers who want a low-power local Gateway. The installation documentation names Raspberry Pi as a deployment option but does not specify a preferred model, supported workload profile, or performance benchmark. Treat it as a host for modest automations, not as a measured replacement for a full server. Use reliable storage and power, keep backups, and move memory-intensive browser or multi-agent jobs to a more capable machine if the workload demands it.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Secure the Gateway before exposing it
Assume every inbound message is untrusted
The project security guidance says: “Treat inbound messages as untrusted input.” A message can contain instructions intended to manipulate an agent, request secrets, or trigger an unsafe tool. Do not grant broad host access merely because a sender is known to you.
Pair unknown direct-message senders
DM-capable channels pair unknown senders by default. Approve a request only after confirming the sender and channel:
Best Value
- 【What you Get】You will get 1*Pi 5 8GB Single Board,1*RasTech Case,1*Active Cooler,1*Screwdriver,1*Installation instructions,12-month free warranty, lifetime service, 24-hour prompt and friendly response.
- 【More Connectors】There are two USB 3.0 ports(5Gbps simultaneously) and two USB 2.0 ports, which triple total bandwidth ,support any combination of up to two cameras or displays. Peak SD card performance is doubled through support for the SDR104 high-speed mode. It provides a smooth desktop experience for you. Offer Gigabit Ethernet and a PCIe interface, along with dual-band Wi-Fi and Bluetooth 5.0/BLE wireless capability. The RasTech Pi 5 Kit use the new 27W 5.1V 5A USB-C power connector.
- 【 Support Dual 4Kp60 Display 】Each of the two microHDMI sockets can control a 4K display at 60 Hertz, now support HDR, offering super HD video for media streaming projects. RPi 5 is the first RPi model that comes with a PCI Express port (PCIe 2.0 x1 with 500 MB/s) to attach SSDs (requires separate M.2 HAT).
- 【 Excellent Chips And Applications】Pi 5 is a full-size Pi computer using silicon built in-house at Pi. The RP1 “southbridge” provides the bulk of the I/O capabilities for Pi 5. Pi 5 is more friendly and convenient in the development of Internet of Things, Web development, machine identification, automatic control and other electronic equipment applications and network.
- 【 Faster CPU, Better GPU 】 Pi 5 features a Broadcom BCM2712 64-bit quad-core Arm Cortex-A76 processor running at 2.4GHz, it delivers a 2–3× increase in CPU performance relative to RaspberryPi 4. The 800MHz VideoCore VII GPU is compatible to OpenGL ES 3.1 and Vulkan 1.2, substantial uplift in graphics performance. Pi 5 Offers lightning-fast CPU speed, a PCI Express interface, a Real Time Clock (RTC) and a power button and runs significantly cooler than Pi 4.
openclaw pairing approve <channel> <code>Do not publish pairing codes in a group or approve requests you cannot attribute.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Configure sandboxing and exposure controls
Main-session tools run on the host unless sandboxing is configured. Read the project’s security, exposure, and sandboxing guidance before connecting other users or exposing the Gateway remotely. Prefer a private network, VPN, or authenticated reverse proxy over an open port. Separate the service account from personal accounts, keep secrets out of prompts and transcripts, and give each agent only the channels, files, browser profiles, and nodes it needs.
Protect high-impact actions
- Require a human approval step for deletion, purchases, account changes, code deployment, or messages sent to large groups.
- Use read-only credentials for reporting jobs.
- Validate webhook signatures and reject unexpected payloads.
- Keep browser sessions dedicated to automation; never reuse a personal session containing saved passwords.
- Back up configuration and transcripts according to your privacy and retention requirements.
Operate, update, and recover the installation
Keep the following command families available for routine administration:
doctorfor diagnosing configuration and environment issues.statusandhealthfor service and runtime checks.logsfor startup failures, channel errors, and automation traces.backupandmigratefor moving or recovering an installation.updatefor applying a newer release.
Before updating, record the current Node.js version, export or back up configuration, and note which channels, plugins, skills, and scheduled jobs are enabled. After updating, run a health check, send a Control UI test message, and test one representative channel and automation. Re-read release notes because versions, defaults, and plugin behavior are volatile.
Troubleshoot the common failure points
The command is missing after installation
Open a new terminal and check node --version and the location of openclaw. If the installer changed your shell profile, reload it. On Windows, verify that the installation directory is in PATH for the same user that runs onboarding.
Onboarding cannot complete a real response
Confirm the provider login or API key, check that Node.js meets the supported range, and try manual provider setup. A saved credential that cannot perform a completion is not usable for automation.
The service is installed but the dashboard is unavailable
Run openclaw gateway status, confirm that the service is running and listening on port 18789, then inspect logs. Check whether another process owns the port and whether the service was installed under a different user. Keep the dashboard local until the service is healthy.
A channel receives messages but does not answer
Verify the channel credential, plugin status, sender pairing, and agent routing. Test the same agent in the Control UI to distinguish a provider problem from a channel problem. For group chats, check that the bot or account has the permissions required by that channel plugin.
An automation runs with too much authority
Stop the job, revoke unnecessary credentials, review transcripts and logs, and narrow its tools, workspace, browser profile, node access, and routing rule. Add an approval gate before re-enabling any destructive step.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Quick Recap
A practical launch checklist
- Confirm Node.js 24.16+ or 26.1+; use Node 26 when possible.
- Prepare a Claude Code or Codex CLI login, or a supported provider API key.
- Install OpenClaw with the platform-appropriate installer or supported package path.
- Run Quick start or Custom setup and verify a real completion.
- Stop the foreground process and run
openclaw gateway install. - Run
openclaw gateway status; confirm port 18789. - Run
openclaw dashboardand send a local test message. - Add one channel with
openclaw channels add; Telegram is the simplest documented first test because it uses a bot token. - Pair only known senders and configure sandboxing before inviting other users.
- Define one small automation with a clear trigger, route, minimal tools, success condition, and destination.
- Back up configuration, test logs and recovery commands, and re-check compatibility after updates.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
