Ollama

Overview
Ollama on OneCloudPlanet is the simplest way to run your own LLM: one command — and the model runs on your GPU instance. Open WebUI is included — a browser chat interface for your whole team.

What is Ollama?
Ollama is the most popular tool for running open models locally: Llama, Qwen, DeepSeek, Mistral, Gemma. One ollama run command downloads and starts a model; the API is OpenAI-compatible.

Why Ollama on OneCloudPlanet?
Privacy: prompts and data never leave your instance.
Recommended configuration: 48 GB GPU (L40S/RTX 6000 Ada) for 7–14B models; A100 80 GB for 32B+. See Cloud GPU.
Hourly billing: an evening of experiments costs a couple of dollars.
Open WebUI included: team chat with zero extra setup.

How to launch
1) Sign in at console.ocplanet.cloud
2) Open Marketplace and choose Ollama
3) Pick a configuration and location (Ukraine, Poland, Netherlands)
4) Wait for deploy and open the web UI
5) Change the default credentials

Post-deploy checklist
• Pull your first model: ollama pull llama3.1
• Close ports with the free Firewall — team access only
• Connect apps via the OpenAI-compatible API
• Add Dify from Marketplace for RAG over documents

FAQ
Which models can I run? Anything from the Ollama catalog: Llama, Qwen, DeepSeek, Mistral, Gemma and hundreds more.
Do I need a GPU? Small models (1–3B) run on CPU instances, but a GPU is recommended for comfortable speed.