For most OpenClaw users, the best VPS is an Ubuntu LTS server with 1–2 vCPUs and at least 2 GB of RAM. Start with 2 GB if the agent will handle ordinary messaging, scheduled jobs, and light tools. Choose 4 GB or more for browser automation, media processing, several channels, or node tools. Keep the Gateway private on loopback and reach it through SSH or Tailscale.
There is no universal provider winner. OpenClaw runs on a wide range of Linux VPS platforms, so the right choice depends on the region your users and services occupy, CPU consistency, backups, network controls, scaling, support, and billing clarity. Treat the VPS as OpenClaw’s source of truth: its state and workspace must be backed up and protected.
As an Amazon Associate I earn from qualifying purchases.
Listed first: Antihost VPS
Antihost sells self-managed Linux VPS on AMD Ryzen 9 9950X3D with NVMe storage and full root access, from 1 to 16 vCPUs and 2 to 64 GB RAM, so the sizes in this guide fit its range. It lists a one-click OpenClaw image, and regions are USA East, the Netherlands and Ukraine. A 14-day free trial is available, so you can test your own workload first. See the Antihost OpenClaw VPS page for the current plans.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Quick answer: which VPS should you choose?
Use this decision table before comparing individual providers. The figures are OpenClaw’s documented operating thresholds, not market benchmarks.
#1 Best Overall
- 12th Intel Alder Lake N95 Processor – The GMKtec G3 S Mini PC is powered by the 12th Gen Intel N95 processor with 4 cores, 4 threads, 6MB cache and a burst frequency up to 3.4GHz. Compared with N100/N5105/N5100/N5095, the N95 delivers up to 36% overall performance improvement. Perfect for routine tasks, office work, and home entertainment, this compact mini desktop is more convenient than traditional bulky PCs.
- 8GB RAM & 256GB SSD Storage – Pre-installed with 8GB DDR4 memory and a fast 256GB M.2 2242 SSD, the G3 S mini desktop offers quicker startup, smoother multitasking, and faster file transfers. Enjoy seamless performance whether you’re working on multiple applications, browsing, or streaming content.
- Rich Interfaces & Connectivity – The G3 S mini computer comes equipped with USB 3.2 (up to 10Gbps), dual HDMI 2.0 (4K@60Hz), and a 3.5mm audio jack. With support for WiFi 5, Bluetooth 5.0, and Gigabit Ethernet (RJ45 1000MbE), it connects easily with monitors, projectors, printers, office equipment, and other peripherals, making it versatile for both home and business use.
- Dual 4K Display Support – Featuring upgraded Intel UHD Graphics (up to 1000MHz), the G3 S supports 4K video playback and AV1 decoding for a smooth viewing experience. With dual HDMI outputs, you can connect two 4K@60Hz displays simultaneously, enabling efficient multitasking for work and entertainment.
- GMKTEC WARRANTY - GMKtec offers a 3-year limited warranty (1 year replacement + 2 years parts replacement) for each mini PC, starting from the date of the purchase effective on all sales starting Oct. 2026. All defects due to design and workmanship are covered. With a professional after sales team always ready to attend to your needs, you can simply relax and enjoy your mini PC
| Situation | Recommended VPS configuration | Why |
|---|---|---|
| Testing or a very light, single-channel setup | 1 vCPU, 1 GB RAM, at least 500 MB disk | This is the documented minimum, with little headroom for logs, media, or concurrent work. |
| Most personal and small-team deployments | 1–2 vCPUs, 2 GB or more RAM, Ubuntu LTS | Provides practical headroom for the Gateway, logs, normal integrations, and more than one channel. |
| Browser automation, media tools, or node tools | 2 or more vCPUs and 4 GB or more RAM | These workloads can consume substantially more memory and CPU than a text-only agent. |
| Growing production deployment | At least 4 GB RAM, reliable backups, and a clear upgrade path | Additional channels, concurrent jobs, and larger workspaces increase resource and recovery requirements. |
OpenClaw’s Linux-server documentation names Azure, Daytona, DigitalOcean, exe.dev, Fly.io, GCP Compute Engine, Hetzner, Northflank, Oracle Cloud, Railway, Render, Raspberry Pi, and Upstash Box as deployment options. It also notes that AWS EC2, Lightsail, and the free tier can work well. Those names identify supported paths, not a current ranking by price, latency, uptime, or support.
How OpenClaw uses a VPS
The Gateway runs continuously on the server and owns OpenClaw’s state and workspace. Your laptop or phone is a client: you connect to the Control UI or use Tailscale or SSH to reach the Gateway. This arrangement keeps the agent available when your personal computer sleeps, reboots, changes networks, or installs updates.
The VPS is the source of truth
Configuration, credentials, logs, generated files, and workspace data live on the VPS. A lost instance can therefore mean lost agent state, not merely a period of downtime. Set up snapshots or another backup method before treating the installation as production. Test that you can restore the state and workspace to a fresh server; an untested backup is only an assumption.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Why a dedicated host is preferable
OpenClaw’s FAQ describes a dedicated host such as a VPS, Mac mini, or Raspberry Pi as always-on, less exposed to sleep and reboot interruptions, and easier to keep under clean permissions. A shared laptop or desktop can work for testing or active use, but sleep, operating-system restarts, and updates can pause the agent at inconvenient times.
Minimum requirements and a sensible target
CPU
The documented floor is 1 vCPU. That can run a very light installation, but CPU demand rises with browser sessions, media handling, concurrent jobs, and node tools. One to two vCPUs is the practical starting point for ordinary use. If jobs regularly overlap or browser automation is central to the workflow, move to a larger plan rather than relying on CPU bursting that may not be consistent.
Memory
OpenClaw documents 1 GB RAM as the absolute minimum and recommends 2 GB or more. The extra memory covers the operating system, Node runtime, Gateway, logs, integrations, and temporary work. Browser automation and node tools are specifically called out as resource hungry; 4 GB or more is a safer baseline when they are enabled.
Storage
The approximate minimum is 500 MB of disk. That number is a floor for a minimal installation, not a comfortable production allocation. Logs, media, browser data, package caches, backups, and workspace files accumulate over time. Check free space periodically and choose a disk size that leaves room for your retention policy and at least one recovery copy stored separately from the running instance.
Rank #2
- [15W Ryzen 7 Agentic PC for Everyday Workflows] Powered by the AMD Ryzen 7 7730U processor (8 Cores, 16 Threads), the GEEKOM A5 is built for sustained productivity. It doubles as your cloud-native Agentic AI assistant, seamlessly hosting cloud AI tasks, automating office workflows, and handling intelligent document summarization without complex local deployment. Smoothly manage Microsoft Office, dozens of browser tabs, heavy Excel spreadsheets, and remote learning throughout your workday.
- [Smart Value Now, Expandable for Tomorrow] Equipped with 16GB RAM and a fast 256GB PCIe NVMe SSD for snappy daily performance, the A5 offers incredible value. Need more space later? It features dual-slot DDR4 RAM (upgradable to 64GB) and supports an M.2 SSD up to 4TB. With an extra M.2 2242 slot and 2.5" HDD bay for up to 10TB total storage, you get the flexibility to scale your storage seamlessly as your needs grow, beating soldered LPDDR solutions.
- [Multi-Display Connectivity for Maximum Productivity] Create a complete workstation with support for up to four displays through Dual HDMI and Dual USB-C ports, including up to 8K output via USB-C. Stay connected with Wi-Fi 6, Bluetooth 5.4, a 2.5GbE LAN port, SD card reader, and multiple USB ports for fast networking, efficient multitasking, and seamless connectivity across all your devices.
- [Built to Stay Cool, Quiet & Reliable] More than fast, the GEEKOM A5 is built to last. A reinforced one-piece all-metal internal frame enhances structural strength, while the upgraded IceBlast 3.0 cooling system improves cooling efficiency by up to 42% with up to 35% greater airflow for quieter operation. Backed by 339 reliability tests and a 72-hour full-load aging test, it's engineered for dependable long-term performance.
- 🏢[Business-Ready, Compact & Efficient] Pre-installed OS, the GEEKOM A5 supports Wake-on-LAN, Scheduled Power On, and Group Policy, making deployment and remote management simple for businesses. Its ultra-compact 0.6L design fits neatly behind monitors or into space-limited workstations while delivering excellent power efficiency for home offices, front desks, and commercial environments.
Operating system
Ubuntu LTS is the best-tested path. A modern Debian or Ubuntu release is also appropriate. Select a current LTS image rather than an end-of-life distribution, and keep security updates enabled. The objective is a predictable package base and an uncomplicated Node installation, not a particular virtualization brand.
Node.js runtime
The installation documentation requires Node 24.16 or newer, or Node 26.1 or newer, and recommends Node 26. Confirm the runtime after provisioning with node --version. If the image ships an older release, install a supported version before installing OpenClaw. Pin or document the version so a later system update does not silently move the Gateway onto an unsupported runtime.
How to compare VPS providers for OpenClaw
Compare the actual region and plan you would buy, not a provider’s headline offering. The following criteria have the greatest effect on an OpenClaw deployment.
1. Region and network path
Choose a data-center region close to the people, APIs, messaging services, and websites the agent uses most. A nearby region can reduce round-trip delay and make interactive Control UI sessions feel more responsive. If your users and services are spread across continents, prioritize the location of the latency-sensitive integrations and verify that the provider offers a suitable region before creating the server.
2. Consistent CPU and RAM allocation
Check whether the plan supplies dedicated or shared resources, what happens during host contention, and how upgrades change the allocation. OpenClaw’s minimum numbers are useful only when the VPS can deliver them reliably. A nominally larger plan that frequently throttles may behave worse than a smaller plan with predictable CPU.
3. Snapshots and backups
Because the Gateway owns state and workspace, look for automated snapshots, retention controls, and a way to copy backups outside the running server. Confirm whether snapshots are crash-consistent or application-consistent, how restoration works, and whether backup storage is billed separately. Keep at least one recovery copy independent of the VPS account or region when the data matters.
4. Administrative controls
The provider should give you normal SSH access, a firewall or security-group layer, private networking where available, and enough control to install Tailscale. These controls let you keep the Gateway private instead of exposing an administration port to the public internet.
Rank #3
- 【AMD Ryzen 3 5300U CPU: Outperforms N150 & 3500U】 BOSGAME E5 mini PC is powered by the TSMC 7nm FinFET architecture AMD Ryzen 3 5300U processor (4 Cores, 8 Threads, up to 3.8GHz boost, 6MB total cache). Compared to low-end Intel N150 or 3500U chips which only have 4 single threads and throttle under load, the 5300U delivers over 30% faster multi-core speed. Run 30+ browser tabs, large Excel sheets, and Zoom meetings simultaneously without system lag.
- 【8GB DDR4 RAM & 256GB NVMe SSD Storage】 Installed with high-speed 8GB DDR4 dual-channel memory and a fast 256GB M.2 2280 SSD, eliminating slow boot times and application loading delays. To accommodate growing data requirements, the upgradeable hardware design features dual SODIMM slots that allow you to expand memory up to 64GB RAM, ensuring smooth operation during heavy multitasking.
- 【High-Capacity Dual M.2 SSD Storage Expansion】 Never worry about running out of space for your business files. In addition to the pre-installed 256GB system drive, the motherboard houses an extra empty internal M.2 2280 NVMe PCIe 3.0 slot. This allows you to easily add a second solid-state drive for up to an additional 2TB of storage capacity (upgrades not included) without needing to remove or reinstall the original operating system.
- 【Radeon 6-Core Graphics & Triple 4K Displays】 Integrated with official AMD Radeon Graphics (6 Graphics Cores, 1500 MHz frequency) for casual gaming, photo editing, and crisp 4K media decoding. Featuring 1x HDMI 2.0 port, 1x DisplayPort, and 1x Full-Function Type-C port, the E5 outputs true 4K@60Hz resolution to three monitors at once. This multi-screen setup eliminates constant window-switching for traders, programmers, and office workers.
- 【Dual 2.5GbE LAN Ports for Advanced Networking】 Experience fast wired network transmission speeds up to 2500Mbps without lagging or buffering. The integration of dual 2.5 Gigabit Ethernet ports (powered by Realtek RTL8125 controller) makes this compact computer an exceptional hardware choice for tech enthusiasts. Easily configure it into software routers, hardware firewalls (pfSense, OpnSense), home NAS servers, or local homelabs.
5. Ubuntu, Debian, and Node support
Verify that the image catalog includes a supported Ubuntu LTS or modern Debian release. Check how quickly security updates and new images appear, and whether you can install Node 26 without replacing provider-managed components. A clean, documented install path is more valuable than a long list of operating-system images.
6. Scaling and recovery
Find out whether you can resize CPU, RAM, and disk without rebuilding the server, and whether the provider offers a rescue console or image export. A clear upgrade path matters when you add channels, browser automation, media tools, or nodes. Also check how quickly a replacement instance can be created from a snapshot.
7. Support and billing transparency
Review support hours, response expectations, incident communication, billing granularity, and charges for snapshots, backups, public IPv4 addresses, and outbound traffic. OpenClaw’s official pages do not publish a current cross-provider price, latency, uptime, or affiliate comparison, so do not treat any universal “cheapest” or “fastest” ranking as authoritative without checking the exact plan and region.
Supported provider options without a false ranking
OpenClaw’s provider picker includes the following choices. Evaluate each against the criteria above; the official list does not establish that one is best for every workload.
| Option named in OpenClaw documentation | What to verify before choosing |
|---|---|
| DigitalOcean | Available region, 2 GB and 4 GB plan details, backup and snapshot retention, and resize procedure. |
| Hetzner | Region near your services, resource consistency, backup pricing, firewall controls, and support terms. |
| AWS EC2 or Lightsail | Instance family, storage and data-transfer charges, security-group setup, and the process for creating repeatable backups. |
| Azure or GCP Compute Engine | Regional availability, billing complexity, startup image, firewall rules, and a practical way to limit administrative exposure. |
| Fly.io, Railway, Render, Northflank, or Upstash Box | How persistent storage, private networking, backups, and long-running Gateway processes are implemented on the selected service. |
| Daytona or exe.dev | Whether the offered Linux environment provides the disk persistence, SSH access, and always-on behavior your Gateway needs. |
| Oracle Cloud | Current capacity in your desired region, support level, network controls, and the terms of any free allocation. |
| Raspberry Pi | Power, connectivity, physical access, and off-device backup; it is self-hosting rather than a conventional cloud VPS. |
For a direct “Hetzner or DigitalOcean?” decision, do not use a blanket winner. Pick the one with the better region for your traffic, predictable resources at the size you need, usable backups, and the administration controls you can operate confidently. Recheck the exact plan because offerings and regions change.
Recommended Free Tools
Provisioning an OpenClaw VPS: a safe sequence
- Select Ubuntu LTS or modern Debian. Create the smallest plan that matches the workload table, but avoid 1 GB RAM when you already know browser automation, media, or multiple channels are required.
- Record recovery details. Save the instance identifier, region, image version, SSH key location, backup policy, and billing settings in a password manager or operations record.
- Secure administration before installing the agent. Apply the provider firewall, use key-based SSH, disable password authentication after confirming key access, and keep the system updated. If you want tailnet-only administration, install Tailscale first.
- Verify a second SSH path over Tailscale. Open a separate session using the server’s Tailscale address and confirm that it works before restricting public SSH. If Tailscale is not suitable, apply equivalent hardening to the public SSH path and keep a tested recovery-console method.
- Install and verify Node. Use Node 26 where possible; the documented alternatives are Node 24.16 or newer and Node 26.1 or newer. Run
node --versionand confirm the result before proceeding. - Install OpenClaw using its current Linux-server instructions. Do not substitute an unverified package or an outdated command. The installation method can change, while the supported runtime requirement remains the version gate to check first.
- Keep the Gateway on loopback. Bind it to the local interface by default. Use an SSH tunnel or Tailscale Serve for the dashboard rather than opening the Gateway directly to the internet.
- Configure authentication for any broader binding. If the Gateway must bind to a LAN or tailnet address, set a shared Gateway authentication token or password unless a trusted proxy is deliberately handling authentication.
- Create the first backup. Back up state and workspace after configuration, then schedule recurring snapshots or file-level backups. Store a copy outside the live instance and periodically perform a restoration test.
- Exercise the workload. Send representative messages, run scheduled jobs, test every channel, and try the browser or node tools you intend to use. Watch memory, CPU, disk, and logs before inviting more users.
Keeping the Gateway private
Loopback is the secure default
OpenClaw’s documented secure default is to keep the Gateway on loopback and access it through an SSH tunnel or Tailscale Serve. This avoids publishing the dashboard or control endpoint as a public service. Provider firewalls are useful, but they are not a replacement for binding the Gateway to the local interface.
When a tailnet address is appropriate
A tailnet can provide private access from approved devices without exposing the Gateway to the public internet. Install Tailscale before restricting public SSH, verify a separate session over the Tailscale address, and keep a recovery path. Device enrollment and account security become part of your threat model, so remove old devices and protect the identity provider used for the tailnet.
Rank #4
- This Certified Refurbished product is tested and certified to look and work like new. The refurbishing process includes functionality testing, basic cleaning, inspection, and repackaging. The product ships with all relevant accessories, a minimum 90-day warranty, and may arrive in a generic box. Only select sellers who maintain a high-performance bar may offer Certified Refurbished products on Amazon.com.
- Dell OptiPlex 7050 Micro Computer, Intel Quad Core i5-6500T up to 3.1GHz, 16G DDR4, 256G SSD.
- Includes: USB Keyboard & Mouse, Microsoft office 30 days free trail.
- Ports: 1 x RJ-45, 1 x HDMI, 1 x DP, 6 x USB 3.0.
- 4K Support: Support 4K (3840x2160) Dual display, makes it easy to connect two monitors at the same time, and you can expand working Windows, mirror content, or expand a single window across multiple monitors.
When you bind beyond loopback
Binding to a LAN or tailnet address expands the set of machines that can reach the Gateway. Configure a shared authentication token or password unless a trusted proxy is explicitly responsible for authentication. Limit firewall rules to the required private networks and avoid assuming that an internal address is automatically safe.
Resource planning by workload
| Workload | Starting point | Upgrade signals |
|---|---|---|
| One channel, text-focused tasks, occasional schedules | 1–2 vCPUs and 2 GB RAM | Memory pressure, swap activity, slow concurrent responses, or growing logs. |
| Several channels and regular media handling | 2 vCPUs and 4 GB RAM | Long queue times, failed media jobs, or sustained CPU saturation. |
| Browser automation with multiple sessions | At least 4 GB RAM; add CPU for concurrency | Browser crashes, out-of-memory kills, or sessions competing with the Gateway. |
| Node tools and mixed automation | 4 GB RAM or more, with tested CPU headroom | Tool timeouts, high load during overlaps, or insufficient temporary disk. |
Measure during the busiest realistic workflow, not while the server is idle. Memory exhaustion is often the first problem on small instances; disk exhaustion follows when media and logs are retained indefinitely. Set alerts for available memory, disk usage, and repeated process restarts.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Troubleshooting common VPS failures
| Symptom | Likely cause | Action |
|---|---|---|
| Gateway stops whenever you close your laptop | The Gateway is running on the laptop instead of the VPS, or the VPS process is not managed as a persistent service. | Run the Gateway on the dedicated host and configure the installation’s supported service method; verify it after disconnecting SSH. |
| Installation refuses to start | Node is below the documented version. | Check node --version; move to Node 26 or a supported Node 24.16+ release, then retry. |
| Browser tasks crash or the server swaps heavily | Insufficient RAM or too many concurrent sessions. | Reduce concurrency, stop unused tools, inspect logs, and resize to 4 GB or more. |
| Control UI is unreachable | The Gateway is correctly bound to loopback but no tunnel or private route is active. | Reconnect through the documented SSH tunnel or Tailscale Serve path; do not expose the port as a quick fix. |
| SSH is locked out after hardening | Public SSH was restricted before the alternate path was tested. | Use the provider’s recovery console or rescue access, restore a tested key or rule, and repeat the change only after a second session works. |
| State disappears after rebuilding | The workspace and Gateway state existed only on the old instance. | Restore from the external backup or snapshot, then document and automate the backup procedure. |
| Disk usage keeps climbing | Logs, media, browser data, or snapshots are retained without limits. | Set retention rules, move backups off-instance, remove stale artifacts, and increase disk only after fixing retention. |
Is a 1 GB VPS enough?
It can be enough for a minimal, low-concurrency test because 1 vCPU and 1 GB RAM are the documented minimums. It is not a comfortable default. The operating system and Gateway leave little room for logs, media, multiple channels, browser processes, or node tools. If the server will be always-on or important, 2 GB is the more practical starting point; choose 4 GB or more when heavier tools are part of the plan.
Can OpenClaw run on Ubuntu?
Yes. Ubuntu LTS is the best-tested Linux installation path, and modern Debian or Ubuntu systems are also suitable. The important checks are a supported Node runtime, a persistent server, secure Gateway access, and backups of state and workspace.
What makes a VPS “best” for OpenClaw?
The best VPS is the one that reliably meets your workload in the right region and gives you a recoverable, private operating environment. A low advertised price cannot compensate for inconsistent CPU, inadequate memory, missing backups, difficult SSH controls, or a region that makes every integration slow. Start with the documented size, validate your real workload, and resize only when measurements show a need.
Final recommendation
Choose an Ubuntu LTS VPS with 1–2 vCPUs and at least 2 GB RAM for a normal OpenClaw deployment. Move to 4 GB or more for browser automation, media tools, several channels, or node tools. Select among the officially named providers by region, resource consistency, backups, SSH and Tailscale compatibility, scaling, support, and transparent billing—not by an unsupported universal ranking. Keep the Gateway on loopback, administer it through SSH or Tailscale, and back up the state and workspace as carefully as any other production data. Antihost VPS is listed first in this article: it offers a one-click OpenClaw image on a self-managed Linux server with root access, and you can look at it on the Antihost VPS for OpenClaw page.
Quick Recap
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.
