OpenClaw’s documented floor is 1 vCPU, 1 GB of RAM, and about 500 MB of disk. That specification can run a minimal Gateway with one lightly used channel, but it leaves almost no room for operating-system activity, logs, updates, or child processes. For ordinary personal use, start with 1–2 vCPUs and at least 2 GB of RAM. Choose more memory when you add browser automation, media handling, several channels, or node tools.
Antihost offers VPS plans with one-click OpenClaw, NVMe storage and a 14-day free trial.
A VPS is the practical 24/7 option because it stays online when your laptop sleeps. The trade-off is that you administer a headless Linux machine, normally through SSH or Tailscale, and you must treat the VPS workspace and backups as the authoritative copy of your OpenClaw setup.
As an Amazon Associate I earn from qualifying purchases.
The two numbers that matter: minimum and recommended capacity
OpenClaw’s 2026 documentation separates an absolute bootable minimum from a more useful operating target. The minimum is intended to prove that the Gateway can run on a very small virtual machine, not to guarantee comfortable performance under concurrent work.
| Profile | CPU | Memory | Disk | What it is suitable for |
|---|---|---|---|---|
| Absolute minimum | 1 vCPU | 1 GB RAM | Approximately 500 MB | A basic Gateway and one low-concurrency channel, with careful resource use |
| Safer starting point | 1–2 vCPUs | At least 2 GB RAM | More than the minimum, with room for logs and workspace data | Routine personal use, normal updates, and occasional simultaneous activity |
| Expanded workload | 2 or more vCPUs | More than 2 GB when memory pressure appears | Size for retained media, logs, and backups | Browser automation, media processing, multiple channels, or node tools |
The first row is a real OpenClaw requirement, not a theoretical figure. It is also a very tight envelope: the Linux distribution, Node runtime, OpenClaw, logging, and any child process all share the same gigabyte. A small amount of swap can prevent an abrupt crash during a brief spike, but swap is not a substitute for physical RAM when a workload is continuously busy.
#1 Best Overall
- WHY CHOOSE G3 ULTRA MINI PC PENTIUM GOLD 7505 - Choose the Intel Pentium Gold 7505 for snappier everyday responsiveness: It delivers up to 30% faster single-core performance than the Ryzen 5 3500U, making office apps and web browsing feel noticeably quicker, while its Intel UHD Graphics (48 EUs) provides 2.4x the GPU performance of the N100 & N150's 24-EU graphics, ensuring smoother 4K streaming and light photo editing.
- 16GB RAM MEMORY & 512GB STORAGE - GMKtec Nucbox G3 Ultra mini computer is prebuilt with 16GB LPDDR4 RAM at 3200 MT/s, you will enjoy a speedier experience with Built-in 512GB M.2 SATA Hard Drive. Our mini desktop pc boots up in seconds, work on multiple browser tabs, software applications and quickly transfers files. There is a primary slot and secondary expansion storage. Primary slot is M.2 2280 PCIE and secondary slot is M.2 2280 SATA.
- RICH INTERFACE - Nucbox pentium mini computer is equipped with 3* USB 3.2 Gen2 ports, up to 10Gbps/S, 1*USB 2.0, HDMI(4K@60Hz)*2, 3.5mm Audio Jack. Supports WiFi 6, and Gigabit Ethernet RJ45 2.5GbE network connectivity, Bluetooth 5.2. This Mini PC supports multiple device connection and can be used with servers, monitoring equipment, office equipment, displays, projectors, televisions, etc.
- 4K DUAL SCREEN DISPLAY - Mini desktop computer is equipped with upgraded Intel Graphics(max 1000MHz), supports 4K video playback and AV1 decoding, connect the pc with a projector as a home theatre, enjoy a variety of entertainments. Two HDMI 2.0 ports allows you to multi-task efficiently on two 4K@60Hz displays.
- UPGRADED COOLING FAN - The G3 Ultra has upgraded the cooling fan to reduce fan noise and thermals. We are using an upgraded thermal paste as well to help reduce heat on the CPU.
How your workload changes the size
Basic Gateway with one channel
If you are testing OpenClaw, handling one channel, and keeping concurrency low, a 1 vCPU/1 GB machine can be enough. Expect to schedule updates outside busy periods and to watch memory rather than assuming the VM has spare capacity. The documented 1–2 vCPU and 2 GB-or-more recommendation is a better baseline if the Gateway will be used every day.
Several channels or media activity
Each active channel can create overlapping work: message handling, authentication, network requests, logging, and sometimes media conversion or temporary files. Use at least 2 GB of RAM for this class of deployment and leave additional headroom when several channels can receive events at the same time. CPU demand rises with simultaneous work, so two vCPUs is a sensible starting point when activity is no longer strictly sequential.
Recommended Free Tools
Browser automation
Browser processes are often the first component to expose a small VM’s limits. A browser can consume substantially more memory than the Gateway itself, especially with multiple pages, extensions, screenshots, downloads, or parallel sessions. Treat 2 GB as a starting point rather than a ceiling, measure peak use, and move to a larger memory tier if the kernel begins reclaiming aggressively or processes are killed.
Node tools and local capabilities
Node tools can launch local helper processes in addition to the Gateway. Their combined footprint depends on what each tool does, how long it runs, and whether several requests overlap. Plan capacity for the Gateway plus the largest child process you expect, not for the Gateway in isolation. If a tool invokes a compiler, browser, media utility, or data conversion step, the peak can be much higher than the idle reading.
How many vCPUs does OpenClaw need?
One vCPU satisfies the documented minimum and is adequate for a lightly used, single-channel Gateway. It does not provide much scheduling room when the operating system, Node, network handling, and a child process all need CPU time together.
Two vCPUs give the routine-use profile more breathing room. They help when messages arrive concurrently, logs are written while a task runs, or a tool performs CPU work in the background. Additional vCPUs are justified by measured parallelism, not by the mere presence of OpenClaw: adding cores will not fix a workload that is blocked on memory or I/O.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #2
- 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
- Choose 1 vCPU for a low-cost, low-concurrency Gateway with one channel.
- Choose 2 vCPUs when you expect overlapping channel activity, media work, or occasional automation.
- Choose more only after observing sustained CPU saturation or a known parallel workload; increase RAM first when the evidence points to memory pressure.
Disk sizing: the 500 MB figure is only the floor
OpenClaw documents approximately 500 MB of disk for the minimum VPS. That figure covers a minimal installation, not unlimited history. The operating system, Node packages, logs, downloaded media, browser caches, temporary files, and backups all consume storage over time.
When comparing plans, check whether the advertised disk is local SSD, network-backed storage, or an add-on volume, and whether snapshots or backups count against the quota. Keep logs on a retention policy, remove abandoned temporary files, and reserve free space for package updates. A VM that reaches 100 percent disk usage can fail even when CPU and RAM look healthy.
Operating system and Node requirements
The best-tested path in the current documentation is Ubuntu LTS or another modern Debian/Ubuntu distribution. OpenClaw can run on any Linux server or cloud VPS, but an up-to-date LTS release reduces surprises with package managers, system services, and security updates.
The current runtime requirement is Node 24.16 or newer, or Node 26.1 or newer; Node 26 is recommended. Verify the installed runtime before troubleshooting OpenClaw itself. A version that is below either supported floor can produce installation or startup errors that look like application failures.
- Provision the VPS with an Ubuntu LTS or comparable modern Debian-based image.
- Install a supported Node release and confirm it with
node --version. Prefer Node 26 unless a project constraint requires the supported Node 24 line. - Install OpenClaw using its current Linux-server instructions, then run the Gateway under a dedicated, non-root account.
- Keep the Gateway bound to loopback unless you have a specific reason to expose it. Reach it through an SSH tunnel or Tailscale Serve rather than opening an administrative interface to the public internet.
- Record the VPS region, image, Node version, OpenClaw configuration, and service settings so the machine can be rebuilt.
Why use a VPS instead of a laptop?
OpenClaw’s documentation describes a VPS as the option for 24/7 reliability. A hosted VM remains powered on during laptop sleep, reboots, travel, and home-network outages. It also gives the Gateway a stable network location and a predictable Linux environment.
The cost is operational responsibility. A VPS is headless: you use SSH, a terminal, and sometimes Tailscale rather than a local desktop. You must apply security updates, protect credentials, monitor disk and memory, and plan recovery if the provider or VM fails. Keep a current backup of the VPS state and OpenClaw workspace because that data is the source of truth for your deployment.
Rank #3
- ➊ [ Trusted Quality for Everyday Agentic AI ] GEEKOM equips its SSDs with reliable original-grade flash and conducts rigorous stability testing to support dependable everyday operation. This commitment to quality is backed by a 3-year warranty. Simply connect the Air12 to cloud AI services for research, writing, study support and daily productivity—no NPU or complex local setup required. Designed for students, home users, light office work and first-time buyers, the Air12 is a high-value Cloud Agentic PC for everyday tasks
- ➋ [ Intel 7505 processor ] Powered by the Intel 7505 processor (2 cores, 4 threads, up to 3.5GHz), the GEEKOM Mini PC Air12 delivers smooth performance for everyday computing, office tasks, and home entertainment. With enhanced single-core processing, it handles daily workloads efficiently and responsively. Compact, quiet, and energy-efficient — a solid alternative to bulky desktops.
- ➌ [440lbs(200kg) Pressure Rated Metal Frame for Demanding Environments] Unlike the Plastic Shells You’ll Find on Most Mini PCs, geekom Mini Air12 features a triple-reinforced ABS+PC shell, precision-crafted metal frame and baseplate—engineered to withstand up to 440 lbs of pressure for the perfect balance of strength and thermal efficiency. Tool-free upgrades, shock-absorbing feet, and a 3D antenna deliver true durability
- ➍ [Dual-Channel RAM & NVMe SSD Expandability] Ships with 8GB DDR4 RAM and a 256GB NVMe SSD for smooth everyday performance. Dual memory slots and dual storage slots give you the flexibility to upgrade to 64GB RAM and 2TB SSD, so your system can adapt as your workload grows. Enjoy faster load times, smoother multitasking, and long-term reliability.
- ➎ [Triple 4K Displays for Maximum Productivity] Connect up to three 4K monitors via HDMI 2.0, Mini DisplayPort 1.4, and USB-C — ideal for stock trading dashboards, multi-tab research, office document editing, and light spreadsheet work. WiFi 6 and Bluetooth with high-gain antenna ensure stable wireless connections throughout your workspace. 5x USB ports and a full-size SD card reader provide quick access to peripherals and camera files — no adapters required.
Choosing a VPS provider without a misleading “best” list
There is no current, authoritative cross-provider price or performance benchmark in the OpenClaw documentation. The official provider picker names DigitalOcean, Hetzner, AWS, Fly.io, GCP, Azure, Oracle Cloud, Railway, Render, and other services. The right choice depends on the plan details available in your region and your tolerance for administration.
| Comparison axis | What to check | Why it matters to OpenClaw |
|---|---|---|
| vCPU and RAM | Dedicated versus shared CPU, burst rules, and upgrade sizes | Determines whether concurrent channels or browser processes have headroom |
| Region | Datacenter distance from users and connected services | Lower latency can improve interactive channel behavior |
| Storage | SSD or network storage, included capacity, and expansion cost | Logs, media, caches, and backups grow beyond the installation floor |
| Backups | Automated snapshots, retention, restore testing, and extra fees | A backup is useful only if it can restore the Gateway and workspace |
| Network limits | Transfer allowance, egress pricing, IPv4 policy, and firewall controls | Media-heavy or tool-heavy workloads can generate more traffic than a text-only Gateway |
| Access | Reliable SSH, console recovery, and compatibility with Tailscale | You need secure headless administration and a way back in after a network mistake |
| Total monthly cost | Compute, storage, backups, bandwidth, and taxes rather than headline VM price | The cheapest instance can become expensive after required add-ons |
Start with a plan that meets the 2 GB recommendation, then resize based on observed peaks. A provider with clear snapshots and an easy vertical upgrade can be more valuable than a slightly cheaper plan that is difficult to recover.
A practical sizing workflow
- List the workload. Count channels, identify whether media is processed, note any browser automation, and list node tools that can run concurrently.
- Set the baseline. For a simple personal Gateway, begin at 1–2 vCPUs and at least 2 GB RAM. Use the 1 GB tier only when low cost and low concurrency are deliberate requirements.
- Reserve storage. Budget for the OS, Node dependencies, logs, media, caches, and at least one restorable backup. Do not size the disk to the approximately 500 MB installation floor.
- Install the supported runtime. Use Node 26 where possible, or Node 24.16+; confirm the version before installing OpenClaw.
- Secure access. Use SSH keys, a firewall, and loopback binding for the Gateway. Add Tailscale Serve or an SSH tunnel when remote access is needed.
- Observe real peaks. During normal use and a representative busy period, inspect memory, swap, CPU load, disk space, and process counts. The peak, not the idle reading, determines whether the plan is adequate.
- Resize deliberately. Increase RAM when the kernel or child processes run out of memory; increase vCPUs when CPU saturation is sustained; increase disk when free space falls toward the provider’s warning threshold.
Diagnosing crashes and slowdowns
Child processes are OOM-killed
An out-of-memory kill means the combined workload exceeded the VM’s available memory. First identify which operation was running and whether several requests overlapped. The documented remedies are to raise the memory limit, reduce concurrency, or apply explicit limits so one child process cannot consume the entire machine.
On a systemd-managed service, a MemoryMax= control can cap a unit. Containers provide an equivalent memory limit. These controls protect the host, but a limit that is too low simply causes the child process to fail sooner, so pair the limit with a workload-appropriate RAM allocation.
Repeated startup work on a small VM
Low-power VMs can spend noticeable CPU time compiling or respawning processes. The Linux guidance documents NODE_COMPILE_CACHE for reusing Node’s compile cache and OPENCLAW_NO_RESPAWN=1 to avoid repeated respawn behavior where that trade-off is appropriate. Apply these settings only with an understanding of how your service is supervised: disabling respawn removes an automatic recovery mechanism.
High CPU with normal memory
Check whether browser pages, media conversion, or a node tool is active before changing providers. If one operation consistently saturates the available vCPU, reduce parallel work or move to two or more vCPUs. More cores will not improve a task that is waiting on a remote service, disk, or memory.
Rank #4
- 【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.
Disk fills unexpectedly
Inspect logs, media directories, caches, and backup snapshots. Set retention rules and verify that cleanup jobs actually run. Expand the volume before updates fail, and keep a tested backup outside the VM.
The Gateway is unreachable
Confirm that the process is running, the VPS firewall allows only the intended management path, and the Gateway is listening on the expected local address. If it is bound to loopback, connect through the configured SSH tunnel or Tailscale Serve; do not expose it publicly just to bypass a tunnel problem.
Example plans by use case
| Use case | Starting allocation | Upgrade trigger |
|---|---|---|
| Evaluation or one quiet channel | 1 vCPU, 1 GB RAM, disk above the documented floor | Any swap thrashing, OOM event, or routine concurrent activity |
| Daily personal use | 1–2 vCPUs, at least 2 GB RAM | Repeated CPU saturation, growing logs/media, or additional channels |
| Several channels with media | 2 vCPUs and more than the minimum memory headroom | Peak memory pressure, media queues, or overlapping jobs |
| Browser automation and node tools | At least the 2 GB starting point, monitored closely | Browser or child-process OOM kills, high swap, or sustained CPU load |
Raspberry Pi and other self-hosted alternatives
The official provider page also lists Raspberry Pi ARM self-hosting. That is a physical, always-on alternative rather than a VPS: you own the hardware, power, storage, and network connection. It can appeal if you want local control and already have reliable electricity and internet service.
Do not infer a particular Raspberry Pi model’s performance from the VPS minimum. ARM board specifications, cooling, storage media, and concurrent workloads differ. Apply the same method: install a supported Linux and Node runtime, measure peak memory and CPU, secure remote access, and maintain backups.
Security and continuity checklist
- Use a current LTS Linux image and apply security updates.
- Administer through SSH keys or an equivalent strong method; disable unnecessary public services.
- Keep the Gateway on loopback unless a documented access path requires otherwise.
- Use Tailscale or an SSH tunnel for private remote access.
- Back up the VPS configuration and OpenClaw workspace, then test restoration.
- Monitor RAM, swap, CPU, disk space, and service logs after upgrades.
- Document your Node version and environment settings, including any compile-cache or respawn choices.
Bottom line
A 1 vCPU, 1 GB VPS can run a minimal OpenClaw Gateway, but it is a constrained edge case. For a dependable personal deployment, choose 1–2 vCPUs and at least 2 GB RAM, use Ubuntu LTS or a comparable modern Debian/Ubuntu system, and install Node 26 (or Node 24.16+). Add memory before adding complexity: browser automation, media, multiple channels, and node tools create child processes that can exceed the Gateway’s baseline. Monitor the real workload, keep the Gateway privately reachable through SSH or Tailscale, and resize when measurements—not guesswork—show pressure.
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.
