---
title: 'Using OpenAI-Compatible APIs'
description: 'Presenton supports OpenAI-compatible APIs, allowing you to connect to any custom LLM backend that follows the OpenAI API format.'
---

### 🌐 Example: Run Presenton with a Custom LLM

```bash
docker run -it --name presenton -p 5000:80 \
  -e LLM="custom" \
  -e CUSTOM_LLM_URL="http://XXXXXXXXXXX/v1" \
  -e CUSTOM_LLM_API_KEY="your_custom_api_key" \
  -e CUSTOM_MODEL="your-model-name" \
  -e CAN_CHANGE_KEYS="false" \
  -v "./user_data:/app/user_data" \
  ghcr.io/presenton/presenton:latest
```

---

### 🔧 Environment Variables for Custom LLM

| Variable             | Description                                                             |
| -------------------- | ----------------------------------------------------------------------- |
| `LLM="custom"`       | Use the `custom` value to enable OpenAI-compatible API support          |
| `CUSTOM_LLM_URL`     | Base URL of your OpenAI-compatible API (e.g. `http://XXXXXXXXXXX/v1`)   |
| `CUSTOM_LLM_API_KEY` | API key used for authorization (`Bearer` header)                        |
| `CUSTOM_MODEL`       | ID of the model to use (as defined by your API provider)                |
| `PEXELS_API_KEY`     | *(Optional)* Used to fetch high-quality images to enhance presentations |
| `CAN_CHANGE_KEYS`    | Set to `false` to hide API keys from the frontend                       |