iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Antihost offers VPS plans with one-click OpenClaw, NVMe storage and a 14-day free trial.
OpenClaw’s documented minimum is 1 vCPU, 1 GB of RAM, and approximately 500 MB of disk. That configuration can run a very small, low-concurrency gateway, but it leaves little room for logs, media, updates, or child processes. For ordinary personal use, start with 1–2 vCPUs and at least 2 GB of RAM. Move higher when you use several channels, browser automation, media tools, or node tools.
This guide explains the resource numbers, how to size a VPS for your workload, which operating system and Node.js versions to use, and what to change when a small virtual machine runs out of memory.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
OpenClaw VPS requirements at a glance
OpenClaw’s official 2026 guidance separates an absolute floor from a practical starting point. The floor is useful for testing or a minimal gateway; the recommendation is intended to provide operational headroom.
| Resource | Documented minimum | Recommended starting point | Why the difference matters |
|---|---|---|---|
| CPU | 1 vCPU | 1–2 vCPUs | Concurrent channel work, media handling, browser tasks, and node processes compete for CPU time. |
| Memory | 1 GB RAM | At least 2 GB RAM | Extra memory absorbs logs, media buffers, multiple channels, and child processes without immediate swapping or OOM kills. |
| Disk | Approximately 500 MB | More than the minimum when retaining logs, media, or backups | The documented figure is an application minimum, not a promise that an entire operating system, logs, and backups will fit comfortably. |
| Node.js | Node 24.16+ or Node 26.1+ | Node 26 | Node 26 is the recommended runtime in the current documentation. |
The official Linux-server guide says the Gateway can run on any Linux server or cloud VPS. The FAQ also presents a VPS as the option for 24/7 reliability. A VPS avoids laptop sleep and gives the Gateway an always-on host, but you remain responsible for operating-system updates, access control, backups, and monitoring.
#1 Best Overall
- 2.80 GHz processor speed ensures efficient operation with consistent reliability
- Intel Xeon 2.80 GHz processor provides enterprise-grade performance with built-in security and remote management capabilities
- Quad-core (4 Core) processor core helps server process data quickly and reliably for maximum productivity
- 1 processors supported for faster processing and improved access to data, optimizing performance under heavy loads
- With 16 GB memory, you can multitask between applications seamlessly, keeping productivity high and response times quick
Is a 1 GB VPS enough for OpenClaw?
Yes, 1 GB of RAM is a real documented minimum, but it should be treated as a constrained configuration rather than a general recommendation.
When 1 GB can work
- A single Gateway instance handles one channel or a very small number of interactions.
- Concurrency is low and jobs do not launch many child processes.
- You are testing, learning, or minimizing the monthly cost.
- You are prepared to monitor memory and remove unnecessary services from the VM.
Why 1 GB becomes uncomfortable
A Linux distribution, Node.js, OpenClaw, logging, and any helper processes all share the same memory pool. Browser automation and node tools can create additional processes outside the Gateway itself. Media handling can also require temporary buffers. On a 1 GB VM, a short burst may be enough to trigger the Linux out-of-memory (OOM) killer, even when the Gateway appears idle most of the time.
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 matchUse 1 GB when the workload is deliberately small and cost is the primary constraint. If the Gateway is expected to run continuously for personal automations, the documented 2 GB-plus recommendation is the safer baseline.
How much RAM should you allocate?
Basic Gateway and one channel
Allocate 1 vCPU and 1 GB RAM only for a minimal, low-concurrency deployment. For routine use, choose 1–2 vCPUs and at least 2 GB RAM so normal logging, updates, and occasional bursts do not consume the entire memory budget.
Multiple channels or media
Start at 2 GB RAM when several channels are active or when messages include images, audio, files, or other media. Each channel can generate overlapping work, and media processing adds memory pressure. Leave headroom rather than sizing to the average idle footprint.
Rank #2
- Model: Dell OptiPlex 7050 Small Form Factor (SFF)
- Processor: Intel Core i7-7700 3.60 GHz
- Memory: 32GB DDR4 Ram
- Storage: 1TB Solid State Drive (SSD) Fast Boot + Storage
- Operating System: Windows 11 Pro (64-bit)
Browser automation
Browser automation is usually the first feature that makes a tiny VM feel undersized. A browser process and its pages can consume substantially more memory than a text-only interaction. Treat 2 GB as a starting point, not a ceiling, and watch memory during the largest realistic automation job.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesNode tools
Node capabilities may launch local-tool processes in addition to the Gateway. The correct allocation depends on which tools run, how long they run, and whether they overlap. If node tools are central to your workflow, provision more than the minimum and test under concurrent use.
How many vCPUs does OpenClaw need?
One vCPU satisfies the documented minimum. It is adequate for a lightly used Gateway where jobs are mostly sequential. Two vCPUs provide useful scheduling headroom when channels, media work, browser automation, and node tools overlap.
| Workload | Starting CPU choice | What to monitor |
|---|---|---|
| Testing or one quiet channel | 1 vCPU | CPU saturation during message handling and updates |
| Several channels or regular media | 1–2 vCPUs | Load average and response latency during simultaneous events |
| Browser automation plus node tools | At least 2 vCPUs as a practical starting point | CPU steal time, browser launch time, and overlapping child processes |
More vCPUs do not compensate for insufficient RAM. A VM with spare CPU but repeated OOM kills still needs a larger memory allocation or a less concurrent workload.
Disk space: why the 500 MB figure is only a floor
OpenClaw documents approximately 500 MB of disk as the application minimum. That number should not be interpreted as the complete capacity needed for a comfortable VPS. The operating system, package cache, logs, media files, workspace data, and backups also consume storage.
Rank #3
- Dell PowerEdge R730xd 24B SFF 2U Server
- 2x Intel Xeon E5-2690 v4 2.6Ghz 14-Core (28-cores Total)
- 128GB DDR4 RAM – 4x 1.2TB 10K SAS 2.5” 12Gb/s
- Dell H730P mini 2GB 12Gb/s RAID
- 2x 750W PSU - 2x 10Gb SFP+ 2x 1Gb (RJ45) NIC
Before choosing a plan, determine whether you will retain large media files, keep long log histories, or create local snapshots. Select storage that leaves room for those items and for normal filesystem overhead. If the provider offers separate backup storage, use it rather than filling the boot volume with multiple copies of the same state.
Operating system and Node.js prerequisites
Choose a mainstream Linux distribution
Ubuntu LTS or another modern Debian/Ubuntu distribution is the best-tested path described in the documentation. A minimal image reduces background services, but do not remove packages that your access, monitoring, or update process requires.
Install a supported Node.js release
The current requirement is Node 24.16 or newer, or Node 26.1 or newer. Node 26 is recommended. Verify the active version after installation and before diagnosing OpenClaw behavior; an older system package can be selected accidentally when multiple Node installations exist.
node --versionThe output should meet one of those version ranges. Keep the runtime managed consistently so an unattended operating-system update does not switch the Gateway to an unsupported binary.
Recommended Free Tools
Provisioning an OpenClaw VPS: a practical sequence
- Pick the resource tier. Use 1 vCPU/1 GB only for a deliberately small test. Choose 1–2 vCPUs and 2 GB or more for normal personal use, and add memory for browser, media, or node-heavy workloads.
- Select a current Linux image. Use Ubuntu LTS or another modern Debian/Ubuntu release, then apply its security updates before installing application software.
- Create an administrative access path. Use SSH keys rather than a password-only login. Restrict the account used for routine work and keep a separate recovery method documented.
- Install and verify Node.js. Confirm Node 24.16+ or Node 26.1+, with Node 26 preferred, using
node --version. - Install OpenClaw using its current Linux instructions. Keep the Gateway’s workspace and configuration in a known location; those files represent the state you must back up.
- Keep the Gateway private by default. A loopback-bound Gateway can be reached through an SSH tunnel or Tailscale Serve instead of exposing an administrative port directly to the public internet.
- Exercise the largest expected workload. Send representative messages, process typical media, and run browser or node tasks concurrently. Observe memory and CPU while those jobs overlap.
- Set up updates and backups. Back up the VPS state and workspace because the VPS is the source of truth for a 24/7 deployment.
Remote access without exposing the Gateway
Headless operation normally means administering the machine over SSH. For private access to a loopback-bound Gateway, an SSH tunnel forwards a local port to the VPS without making the service publicly reachable. Tailscale Serve is another documented approach for reaching the service through a private network path.
Rank #4
- MODEL P74439-005: Compact and affordable HPE ProLiant MicroServer Gen11 powered by Intel Pentium Gold G7400 3.7GHz processor, ideal for file sharing, NAS, and basic business workloads
- READY OUT OF THE BOX: Includes 16GB DDR5 UDIMM memory (expandable to 128GB), one 1TB SATA 6G Business Critical HDD, embedded Intel VROC SATA, dedicated iLO-M.2 port kit, 180w external power adapter and 1/1/1 warranty for dependable plug-and-play server operation
- WHISPER-QUIET & SPACE-SAVING: Ultra-compact mini tower design fits easily in small office spaces; supports wall, flat, or vertical placement for deployment flexibility
- INTEGRATED REMOTE MANAGEMENT: Comes with HPE iLO 6 and embedded TPM 2.0 for secure, license-free remote server administration through shared port access
- EXPANDABLE DESIGN: Two PCIe slots (including PCIe 5.0) and four LFF-NHP drive bays provide robust options for storage and component scalability. Features new MR408i-p controller support for enhanced storage performance
Whichever method you choose, protect SSH keys, limit inbound firewall rules, and retain a recovery path in case a networking change locks you out. A VPS improves uptime only when you can reliably maintain and recover it.
Diagnosing memory pressure and OOM kills
Repeated child-process termination is a sizing or concurrency signal. Check system memory while reproducing the failure rather than relying on an idle reading.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →free -h
nproc
uptimefree -h shows available memory and swap, nproc reports visible CPU threads, and uptime provides load information. Correlate those readings with the time a browser, media, or node task starts.
Increase the memory limit
Move from 1 GB to at least 2 GB when ordinary work regularly consumes the available memory. If browser automation or several tools run together, increase memory again rather than trying to hide repeated OOM events with restarts.
Reduce concurrency
Limit simultaneous channel jobs, browser pages, or node tasks when a larger VM is not possible. Queueing work trades speed for stability and can keep a small instance usable.
Best Value
- 【AMD Ryzen 4300U True 4-Core CPU: Outperforms N95 & i3-10110U】KAMRUI P2 Mini PC is equipped with true 4-core AMD Ryzen 4300U processor built on advanced 7nm Zen2 architecture，This means you get consistent, unthrottled performance for hours on end, whether you’re running multiple browser tabs, streaming 4K content, or managing virtual machines. Compare that to Intel N95 (4 efficiency cores that throttle under load) or Intel i3-10110U (only 2 cores total), and the difference is night and day: The KAMRUI P2 AMD Ryzen 4300U (28W) is 40% faster than the Intel i3-10110U and 25% faster than the Intel N95 in multi-core tasks, ensuring smooth, lag-free performance even during heavy workloads.
- 【Integrated AMD Radeon Graphics: 2.5X Stronger for Tri 4K】The KAMRUI P2 AMD 4300U Mini PC have unlocked the full potential of the built-in AMD Radeon Vega 5 graphics with 28W power delivery, making it 2.5 times stronger than the Intel UHD graphics found in the N95 and i3-10110U. This means you can enjoy Tri 4K@60Hz displays without a single stutter, perfect for productivity setups, home theaters, or even light photo/video editing and casual gaming. While the Intel N95/i3-10110U struggle to run a single 4K display without lag, The KAMRUI AMD 4300U Mini PC handles Tri 4K effortlessly, turning your workspace into a high-efficiency hub or your living room into a premium entertainment center.
- 【Large Storage Capacity, Easy Expansion】KAMRUI Pinova P2 mini computers is equipped with 16GB LPDDR4 for faster multitasking and smooth application switching. 512GB M.2 SSD ensures fast startup, fast file transfers and plenty of storage space,eliminating slow loading times and ensuring fast responsiveness. the two storage slots (1x M.2 2280 SATA/NVMe PCIe3.0 slot, 1x M.2 2280 SATA slot) can be combined to provide up to 4TB of total storage(Not included). This gives you enough space for all your projects, media and data.
- 【4K Triple Display】KAMRUI Pinova P2 4300U mini desktop computers is equipped with HDMI2.0 ×1 +DP1.4 ×1+USB3.2 Gen2 Type-C ×1 interfaces for faster transmission, Triple 4K@60Hz Display, KAMRUI P2 mini computer is ideal for visual home entertainment, home office, conference rooms, etc. USB3.2 Gen2 Type-A port ×2 with a transfer speed of up to 10 Gbps (21 times faster than USB 2.0) for efficient data transfer. Ideal for seamless multitasking between spreadsheets, browsers and presentations, or for an immersive entertainment experience.
- 【USB3.2 Gen2 Type-C 10Gbps, Versatile connectivity】KAMRUI P2 mini desktop pc fast and versatile connectivity! The USB3.2 Gen2 Type-C port offers a data transfer rate of 10Gbps and simultaneously supports DisplayPort 1.4 video output. The P2 AMD Ryzen 4300U Mini PC is complemented by Gigabit LAN, WiFi and Bluetooth, so nothing stands in the way of a productive working environment.
Apply explicit service or container limits
The Linux guidance describes systemd’s MemoryMax= and container memory limits as controls for child processes. A limit prevents one workload from consuming the entire host, but setting it too low causes that workload to be killed sooner. Raise the limit when it is the bottleneck; reduce concurrency when the host itself is undersized.
Improve repeated startup behavior on small VMs
For low-power machines that repeatedly restart, the documentation describes NODE_COMPILE_CACHE and OPENCLAW_NO_RESPAWN=1 as tuning options. Apply them only when you understand the operational trade-off: startup behavior may improve, but disabling automatic respawn changes how failures are handled and therefore increases the importance of service monitoring.
How to compare VPS offers for OpenClaw
Do not choose solely by the advertised CPU count. Compare the following characteristics for the region where the Gateway and its users are located.
| Comparison axis | Questions to ask |
|---|---|
| vCPU and RAM | Are the resources dedicated, shared, or burstable? Is the plan easy to resize when browser or media workloads grow? |
| Region | Is a nearby data center available, and does its latency suit your channels and remote administration? |
| Storage | What storage type is included, and how are capacity and I/O limits measured? |
| Backups | Are backups automatic, how long are they retained, and can you restore individual files or only the whole VM? |
| Network limits | What bandwidth, transfer, ingress, and egress limits apply to media-heavy use? |
| Access features | Can you use SSH keys, private networking, firewall rules, and a Tailscale-based access pattern? |
| Total monthly cost | Does the stated price include storage, backups, public IPv4, and overage charges? |
The official provider picker lists services including DigitalOcean, Hetzner, AWS, Fly.io, Google Cloud, Azure, Oracle Cloud, Railway, and Render, among others. The official OpenClaw material does not publish a current cross-provider price or performance benchmark table, so compare the live plan terms yourself rather than treating one provider as universally best.
VPS versus a physical Raspberry Pi
The same official provider information lists “Raspberry Pi ARM self-hosted” as a physical alternative. A Pi can suit readers who prefer to own an always-on host and manage it on their own network. It is not a VPS, and the listing does not establish performance for a particular Raspberry Pi model.
A VPS generally offers data-center power, remote console options, and simpler internet reachability. A self-hosted Pi avoids recurring hosting charges but requires your own power, network configuration, hardware replacement plan, and backup strategy. In both cases, the Gateway’s workspace and configuration need independent backups.
Quick Recap
A sizing decision you can apply immediately
- Testing a single quiet channel: 1 vCPU, 1 GB RAM, and approximately 500 MB is the documented floor; monitor closely.
- Normal personal automation: 1–2 vCPUs and at least 2 GB RAM is the safer starting point.
- Several channels or media: retain at least 2 GB and leave additional headroom for overlapping work.
- Browser automation or node tools: begin at 2 GB or more, test the heaviest concurrent job, and scale memory when monitoring shows pressure.
- Repeated OOM kills: increase RAM, reduce concurrency, or adjust systemd/container limits; do not assume a CPU upgrade alone will fix it.
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.
