ProductFeaturesSelf-Hosted AIHow it worksPricingDocsBlog
Sign inDownload
Self-Hosted AI / Local LLM runtime

Ollama

Run LLaMA, Mistral, Gemma, and Qwen on your own hardware - no API key, no per-token bill.

Download DeployOSSee all AI templates

What is Ollama?

Ollama packages open-weight language models - Llama 3, Mistral, Gemma, Qwen, DeepSeek, and dozens more - behind a simple local API, so a laptop-friendly model runs the same way whether you're calling it from a script, a chat UI, or an agent framework.

It's the piece almost every self-hosted AI stack sits on top of: Open WebUI, LibreChat, Flowise, and Langflow all talk to Ollama the same way they'd talk to OpenAI's API, just pointed at your own server instead.

Running it yourself means the prompts, the documents, and the conversation history never leave your infrastructure - and there's no usage meter running while you experiment.

How DeployOS deploys it

  1. 1DeployOS pulls the official ollama/ollama image and mounts a persistent volume at /root/.ollama, so models you pull stay put across restarts and redeploys.
  2. 2It's exposed on port 11434 over your server's private Docker network - not directly on the internet - so other containers (Open WebUI, Flowise, your own app) can reach it by service name without any manual networking.
  3. 3Once it's running, pull a model from inside the container (or via its API) and it's immediately available to anything else on the same server.
Imageollama/ollama
Port11434
Persistent volume/root/.ollama

Pairs well with

Open WebUIFlowiseLibreChatQdrant

Common questions

What hardware do I need to run Ollama well?

Smaller models (7B-8B parameters, quantized) run acceptably on a 4-8GB RAM VPS with CPU-only inference - slow but usable for personal use. For real speed you want a GPU-backed server; DeployOS doesn't require one, but Ollama will use it automatically if the host has an NVIDIA GPU and the right drivers.

Do I need Open WebUI too?

No - Ollama alone gives you an API you can call directly. Open WebUI just adds a ChatGPT-style browser interface on top, which is what most people actually want day to day.

Is this the same as paying for ChatGPT Plus?

Not quite the same models or polish, but functionally similar for a lot of use cases: your own chat interface, talking to a model that runs entirely on hardware you control, for the cost of the VPS instead of a monthly subscription.

Deploy Ollama on your own server in one click.

Free download, no account needed. AI templates are on paid plans.

Download DeployOS