Skip to main content

​
TypeScript SDK Quickstart

This guide walks you through creating a stream with all parameters configured upfront.

​
Prerequisites

​
Create a Stream

Create a stream with your model, prompt, ControlNets, and IP Adapter all configured: 
import { Daydream } from "@daydreamlive/sdk";

const daydream = new Daydream({
  bearer: process.env.DAYDREAM_API_KEY,
});

const stream = await daydream.streams.create({
  pipeline: "streamdiffusion",
  params: {
    modelId: "stabilityai/sdxl-turbo",
    prompt: "anime character, vibrant colors",
    negativePrompt: "blurry, low quality",
    guidanceScale: 1.0,
    width: 512,
    height: 512,
    controlnets: [
      {
        modelId: "xinsir/controlnet-depth-sdxl-1.0",
        preprocessor: "depth_tensorrt",
        conditioningScale: 0.5,
        enabled: true,
      },
    ],
    ipAdapter: {
      enabled: true,
      scale: 0.5,
    },
    ipAdapterStyleImageUrl: "https://example.com/style.png",
  },
});

console.log("Stream ID:", stream.id);
console.log("WHIP URL:", stream.whipUrl);
console.log("Watch at: https://lvpr.tv/?v=" + stream.outputPlaybackId);
The response includes:
  • id - Unique stream identifier
  • whipUrl - URL to send video via WebRTC WHIP
  • outputPlaybackId - ID to watch the processed output

​
Start Broadcasting

Use the whipUrl with the Browser SDK to send video.

​
Update Prompt

Change the prompt in real-time without reloading the pipeline: 
await daydream.streams.update(stream.id, {
  pipeline: "streamdiffusion",
  params: {
    prompt: "cyberpunk character, neon lights",
  },
});
Hot-swappable parameters (no reload): prompt, negativePrompt, guidanceScale, seed, delta, controlnets[].conditioningScale, ipAdapter.scale

​
Get Stream Status

const status = await daydream.streams.getById(stream.id);
console.log("Stream:", status);

​
List All Streams

const streams = await daydream.streams.getAll();
for (const s of streams) {
  console.log(s.id, s.name);
}

​
Delete a Stream

await daydream.streams.delete(stream.id);

​
Available ControlNets

SDXL (stabilityai/sdxl-turbo):
  • xinsir/controlnet-depth-sdxl-1.0 - Depth guidance
  • xinsir/controlnet-canny-sdxl-1.0 - Edge detection
  • xinsir/controlnet-tile-sdxl-1.0 - Texture preservation
SD1.5 (Lykon/dreamshaper-8, prompthero/openjourney-v4):
  • lllyasviel/control_v11f1p_sd15_depth - Depth
  • lllyasviel/control_v11p_sd15_canny - Canny edges
  • lllyasviel/control_v11f1e_sd15_tile - Tile

​
Next Steps