> ## Documentation Index
> Fetch the complete documentation index at: https://docs.presenton.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Using Ollama Models

> Follow these steps to generate presentations using Ollama on Presenton

## 🔌 Run Presenton with an Ollama Model (Fully Offline)

Presenton supports fully offline operation 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.

### 🚀 Example: Run Presenton with Ollama

```bash theme={null}
docker run -it --name presenton -p 5000:80 \
  -e LLM="ollama" \
  -e OLLAMA_MODEL="llama3.2:3b" \
  -e IMAGE_PROVIDER="pexels" \
  -e PEXELS_API_KEY="your_pexels_api_key" \
  -e CAN_CHANGE_KEYS="false" \
  -v "./app_data:/app_data" \
  ghcr.io/presenton/presenton:latest
```

### 🚀 Example: Run Presenton with your own Ollama server

```bash theme={null}
docker run -it --name presenton -p 5000:80 \
  -e LLM="ollama" \
  -e OLLAMA_MODEL="llama3.2:3b" \
  -e OLLAMA_URL="http://XXXXXXXXXXXXX" \
  -e IMAGE_PROVIDER="pexels" \
  -e PEXELS_API_KEY="your_pexels_api_key" \
  -e CAN_CHANGE_KEYS="false" \
  -v "./app_data:/app_data" \
  ghcr.io/presenton/presenton:latest
```

### 🧾 Ollama Environment Variables

* **`LLM="ollama"`** Select Ollama as the LLM backend.
* **`OLLAMA_MODEL`** Required. The Ollama model to use (e.g., `llama3.2:3b`, `mistral`, `phi3`, etc.). *Example:*

  ```bash theme={null}
  OLLAMA_MODEL="llama3.2:3b"
  ```
* **`OLLAMA_URL`** Optional. Set this if you're running Ollama outside Docker or on a custom host. *Example:*

  ```bash theme={null}
  OLLAMA_URL="http://XXXXXXXXXXXX"
  ```

> You can get a free API key at [https://www.pexels.com/api/](https://www.pexels.com/api/)

> ✅ Add `--gpus=all` to enable GPU acceleration (see [Using GPU](/docs/configurations/using-gpu)).

### 🧠 Supported Ollama Models

| Model               | Size   |
| ------------------- | ------ |
| **Llama Models**    |        |
| `llama3:8b`         | 4.7 GB |
| `llama3:70b`        | 40 GB  |
| `llama3.1:8b`       | 4.9 GB |
| `llama3.1:70b`      | 43 GB  |
| `llama3.1:405b`     | 243 GB |
| `llama3.2:1b`       | 1.3 GB |
| `llama3.2:3b`       | 2 GB   |
| `llama3.3:70b`      | 43 GB  |
| `llama4:16x17b`     | 67 GB  |
| `llama4:128x17b`    | 245 GB |
| **Gemma Models**    |        |
| `gemma3:1b`         | 815 MB |
| `gemma3:4b`         | 3.3 GB |
| `gemma3:12b`        | 8.1 GB |
| `gemma3:27b`        | 17 GB  |
| **DeepSeek Models** |        |
| `deepseek-r1:1.5b`  | 1.1 GB |
| `deepseek-r1:7b`    | 4.7 GB |
| `deepseek-r1:8b`    | 5.2 GB |
| `deepseek-r1:14b`   | 9 GB   |
| `deepseek-r1:32b`   | 20 GB  |
| `deepseek-r1:70b`   | 43 GB  |
| `deepseek-r1:671b`  | 404 GB |
| **Qwen Models**     |        |
| `qwen3:0.6b`        | 523 MB |
| `qwen3:1.7b`        | 1.4 GB |
| `qwen3:4b`          | 2.6 GB |
| `qwen3:8b`          | 5.2 GB |
| `qwen3:14b`         | 9.3 GB |
| `qwen3:30b`         | 19 GB  |
| `qwen3:32b`         | 20 GB  |
| `qwen3:235b`        | 142 GB |

### 📌 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` full image generation functionality.