> ## Documentation Index
> Fetch the complete documentation index at: https://docs.daydream.live/llms.txt
> Use this file to discover all available pages before exploring further.
# Server API
> REST and WebRTC API for real-time video generation
# Scope Server API
The Scope Server provides a REST API for pipeline management and a WebRTC API for real-time video streaming. Build custom applications that generate video in real-time using text prompts or input video.
## What You Can Build
* **Text-to-Video streaming** - Generate live video from text prompts
* **Video-to-Video transformation** - Transform input video in real-time
* **Custom creative tools** - Build VJ software, live performance tools, interactive installations
* **Integrations** - Connect Scope to your existing applications and workflows
When the server is running locally, interactive API documentation (Swagger UI) is available at `http://localhost:8000/docs`. A full OpenAPI reference will be added to these docs soon.
***
## Quick Start
Get streaming video from the Scope API in under 5 minutes. You'll create an `index.html` file that streams real-time generated video using the `longlive` pipeline.
```bash theme={null}
uv run daydream-scope
```
The server runs on `http://localhost:8000` by default.
If not already downloaded:
```bash theme={null}
uv run download_models --pipeline longlive
```
Create a file named `index.html` with the following content:
```html theme={null}
Scope API Quick Start
Connecting...
```
Open `index.html` in your browser. The page will:
* Load the `longlive` pipeline
* Establish a WebRTC connection
* Display real-time generated video based on the prompt
***
## Configuration
### Server Options
```bash theme={null}
uv run daydream-scope [OPTIONS]
Options:
--host HOST Host to bind to (default: 0.0.0.0)
--port PORT Port to bind to (default: 8000)
--reload Enable auto-reload for development
-N, --no-browser Don't open browser automatically
--version Show version and exit
```
### Environment Variables
| Variable | Description |
| :---------------- | :------------------------------------------------------------ |
| `PIPELINE` | Default pipeline to pre-warm on startup |
| `HF_TOKEN` | Hugging Face token for downloading models and Cloudflare TURN |
| `VERBOSE_LOGGING` | Enable verbose logging for debugging |
### TURN Server Credentials
For WebRTC connections that need to traverse firewalls (NAT traversal), TURN servers are used. The server automatically configures TURN credentials when environment variables are set:
**Using Cloudflare (via Hugging Face)**:
```bash theme={null}
export HF_TOKEN=your_huggingface_token
```
If no TURN credentials are configured, the server falls back to Google's public STUN server, which works for direct connections but may not work behind strict firewalls.
***
## API Endpoints
### Health & Info
| Endpoint | Method | Description |
| :---------------------- | :----- | :------------------------------------------- |
| `/health` | GET | Health check |
| `/api/v1/hardware/info` | GET | Get hardware info (VRAM, Spout availability) |
| `/docs` | GET | Interactive API documentation (Swagger UI) |
### Pipeline Management
| Endpoint | Method | Description |
| :-------------------------- | :----- | :-------------------------------------- |
| `/api/v1/pipeline/load` | POST | Load a pipeline |
| `/api/v1/pipeline/status` | GET | Get current pipeline status |
| `/api/v1/pipelines/schemas` | GET | Get schemas for all available pipelines |
### Model Management
| Endpoint | Method | Description |
| :------------------------ | :----- | :-------------------------------------------- |
| `/api/v1/models/status` | GET | Check if models are downloaded for a pipeline |
| `/api/v1/models/download` | POST | Start downloading models for a pipeline |
### WebRTC
| Endpoint | Method | Description |
| :---------------------------------- | :----- | :-------------------------------- |
| `/api/v1/webrtc/ice-servers` | GET | Get ICE server configuration |
| `/api/v1/webrtc/offer` | POST | Send WebRTC offer, receive answer |
| `/api/v1/webrtc/offer/{session_id}` | PATCH | Add ICE candidates (Trickle ICE) |
### Assets
| Endpoint | Method | Description |
| :---------------------- | :----- | :------------------------------------ |
| `/api/v1/assets` | GET | List available assets (images/videos) |
| `/api/v1/assets` | POST | Upload an asset |
| `/api/v1/assets/{path}` | GET | Serve an asset file |
### LoRA
| Endpoint | Method | Description |
| :------------------ | :----- | :------------------------ |
| `/api/v1/lora/list` | GET | List available LoRA files |
***
## Next Steps
Learn how to use the API for specific workflows:
Load and configure pipelines with custom parameters, LoRAs, and VAE types
Update prompts, transitions, and settings in real-time during streaming
Set up WebRTC to receive generated video from text prompts
Transform input video in real-time with bidirectional WebRTC
Guide generation with reference images and control videos