AI & MLOfficial★ 8.1k
Ollama — Run AI Models 100% Locally
Run open-source LLMs entirely on your hardware. Llama 3, Mistral, Gemma, Phi — no cloud API, no data leaving your machine, no monthly bill.
Complete Privacy, Zero Cost
Ollama makes running local AI models as easy as `ollama run llama3`. Combined with OpenClaw, you get a fully autonomous AI agent that never sends a single byte to the cloud.
- Privacy First: All inference happens on your hardware. Perfect for sensitive data.
- No Subscriptions: After hardware cost, running is free forever.
- Model Library: Llama 3.1, Mistral, Gemma 2, Phi-3, CodeLlama, and 100+ more.
- GPU Acceleration: Automatic CUDA/Metal/ROCm detection for fast inference.
Configuration
// openclaw.json
{
"llm": {
"provider": "ollama",
"model": "llama3.1:70b",
"baseUrl": "http://localhost:11434"
}
}Recommended Models
| Model | RAM | Best For |
|---|---|---|
| llama3.1:8b | 8 GB | General tasks |
| llama3.1:70b | 48 GB | Complex reasoning |
| mistral:7b | 6 GB | Fast responses |
| codellama:34b | 24 GB | Code generation |
</ Add to your README
If you use this skill in your project, add the badge to your GitHub README:
[](https://clawdbot.space/en/skills/ollama-provider?ref=badge&utm_source=ollama-provider&utm_medium=readme&utm_campaign=skill-page)