AI agent hosting
A server of your own for AI agents
Launching soon: OpenClaw, Hermes Agent, Claude Code, Ollama and n8n on a cloud server of your own, with root access. You can run them on Cloud VPS today.
Apps
What you can run
Each app with the memory it needs, so you know the server size before plans go on sale.
OpenClaw
A personal AI agent that stays on, reachable from the chat apps you already use, such as Telegram, Discord and Slack.
4 GB of memory or more
Bring your own AI model key or plan.
Its own guide: OpenClaw (opens in a new tab)Hermes Agent
An open-source AI agent by Nous Research that learns skills from its own work and keeps them between sessions.
2 GB of memory or more
Bring your own AI model key or plan.
Its own guide: Hermes Agent (opens in a new tab)Claude Code
Anthropic’s coding agent in a terminal on a server of your own, so long jobs keep running when your laptop sleeps.
4 GB of memory or more
Needs your own Claude Pro, Max, Ultra, Team or Enterprise plan, or a Console account.
Its own guide: Claude Code (opens in a new tab)Ollama + Open WebUI
Open models such as Llama, Qwen, Gemma and DeepSeek on your own server, with a chat page in your browser.
4 GB of memory or more
No API key needed: the models run on the server.
Its own guide: Ollama + Open WebUI (opens in a new tab)n8n on your own server
Workflow automation with AI steps, running on a server of your own rather than a shared cloud.
2 GB of memory or more
n8n’s licence covers your own business’s workflows, not hosting n8n for others.
Its own guide: n8n on your own server (opens in a new tab)
More apps for the same servers: Flowise, AnythingLLM, Langflow, Coolify, Portainer, Uptime Kuma, WordPress and Docker.
Ollama
Which model fits which memory
Models run on the server’s CPU, with no GPU, so its memory decides which model sizes it can run. Smaller models answer faster.
The 7B, 13B and 33B figures are from Ollama’s own guide (opens in a new tab); 4 GB is our Ollama app’s own minimum.
How it works
Your server, your agent
A server of your own
Root access on a cloud server with the memory, storage and transfer its plan lists. Your agent keeps running when your laptop is closed.
Your model, your key
OpenClaw, Hermes Agent and Claude Code work with the AI model key or plan you bring. With Ollama, the models run on the server itself.
Same price at renewal
Each plan renews at the price you sign up at, on the same term. Any change needs 30 days’ notice.
People who answer
Engineers answer live chat every day, with limited hours on Sundays, usually in about 10–15 minutes.
FAQ
Questions, answered honestly
For a business
Running agents for your business?
Tell our engineers what your agent does and who uses it. They will say what it takes to run it on a server of your own.