73 lines
No EOL
3.1 KiB
Text
73 lines
No EOL
3.1 KiB
Text
---
|
|
title: "Using Ollama Models"
|
|
description: "Follow these steps to generate presentations using Ollama"
|
|
---
|
|
|
|
Presenton supports running fully offline using open-source models via [Ollama](https://ollama.com/). This allows you to generate presentations without relying on cloud APIs — keeping your data private and costs low.
|
|
|
|
### 🚀 Run Presenton with an Ollama Model
|
|
|
|
Make sure you have [Ollama installed](https://ollama.com/download) and models downloaded if running them outside Docker.
|
|
|
|
To run Presenton with an Ollama model:
|
|
|
|
```bash
|
|
docker run -it --name presenton -p 5000:80 \
|
|
-e LLM="ollama" \
|
|
-e OLLAMA_MODEL="llama3.2:3b" \
|
|
-e PEXELS_API_KEY="your_pexels_api_key" \
|
|
-e CAN_CHANGE_KEYS="false" \
|
|
-v "./user_data:/app/user_data" \
|
|
ghcr.io/presenton/presenton:v0.3.0-beta
|
|
```
|
|
|
|
> 💡 **Note:** A valid **Pexels API key is required** for image generation when using Ollama models.
|
|
> You can get a free API key at https://www.pexels.com/api/
|
|
|
|
> ✅ Add `--gpus=all` to enable GPU acceleration (see [Using GPU](/docs/configurations/using-gpu)).
|
|
|
|
### 🧠 Supported Ollama Models
|
|
|
|
| Model | Size | Graph Support |
|
|
| ------------------- | ------ | ------------- |
|
|
| **Llama Models** | | |
|
|
| `llama3:8b` | 4.7 GB | ❌ No |
|
|
| `llama3:70b` | 40 GB | ✅ Yes |
|
|
| `llama3.1:8b` | 4.9 GB | ❌ No |
|
|
| `llama3.1:70b` | 43 GB | ✅ Yes |
|
|
| `llama3.1:405b` | 243 GB | ✅ Yes |
|
|
| `llama3.2:1b` | 1.3 GB | ❌ No |
|
|
| `llama3.2:3b` | 2 GB | ❌ No |
|
|
| `llama3.3:70b` | 43 GB | ✅ Yes |
|
|
| `llama4:16x17b` | 67 GB | ✅ Yes |
|
|
| `llama4:128x17b` | 245 GB | ✅ Yes |
|
|
| **Gemma Models** | | |
|
|
| `gemma3:1b` | 815 MB | ❌ No |
|
|
| `gemma3:4b` | 3.3 GB | ❌ No |
|
|
| `gemma3:12b` | 8.1 GB | ❌ No |
|
|
| `gemma3:27b` | 17 GB | ✅ Yes |
|
|
| **DeepSeek Models** | | |
|
|
| `deepseek-r1:1.5b` | 1.1 GB | ❌ No |
|
|
| `deepseek-r1:7b` | 4.7 GB | ❌ No |
|
|
| `deepseek-r1:8b` | 5.2 GB | ❌ No |
|
|
| `deepseek-r1:14b` | 9 GB | ❌ No |
|
|
| `deepseek-r1:32b` | 20 GB | ✅ Yes |
|
|
| `deepseek-r1:70b` | 43 GB | ✅ Yes |
|
|
| `deepseek-r1:671b` | 404 GB | ✅ Yes |
|
|
| **Qwen Models** | | |
|
|
| `qwen3:0.6b` | 523 MB | ❌ No |
|
|
| `qwen3:1.7b` | 1.4 GB | ❌ No |
|
|
| `qwen3:4b` | 2.6 GB | ❌ No |
|
|
| `qwen3:8b` | 5.2 GB | ❌ No |
|
|
| `qwen3:14b` | 9.3 GB | ❌ No |
|
|
| `qwen3:30b` | 19 GB | ✅ Yes |
|
|
| `qwen3:32b` | 20 GB | ✅ Yes |
|
|
| `qwen3:235b` | 142 GB | ✅ Yes |
|
|
|
|
> ✅ **Graph Support** means the model can generate charts and diagrams in presentations.
|
|
|
|
### 📌 Additional Notes
|
|
|
|
- Use the `OLLAMA_MODEL` environment variable to select any supported model.
|
|
- Ensure your system has enough RAM or GPU memory to handle the model.
|
|
- Always include a `PEXELS_API_KEY` for full image generation functionality. |