SEQUENCER
Tuning into this reality.

Connecting to the world and its live audience.

AI WORLDSTREAM API
EARLY ACCESS
Generate a world. Keep it streaming.
Send a prompt and image, receive a continuous web video stream, then change the world at any time through the API.
LIVE OUTPUT
world_ws_7f2c
LATEST API UPDATE
200 OK
“Follow the glowing river into a city waking at sunrise.”
Generating next scene
Prompt + image input
Stable playback URL
Live API updates
Continuous generation
A video session you can keep directing.
Worldstreams stay open to new creative input. Your application can update the story, setting, characters, and visual direction while viewers remain connected.
One continuous stream
Create a session once and keep a stable playback URL as new video is generated.
Live prompt control
Send new direction while the stream is running and shape what the world creates next.
Image updates
Introduce a new reference image, character, setting, or visual direction through the API.
Web-ready playback
Put the returned stream into a website, app, game, installation, or live event.
One session. Continuous control.
Create the stream, give its playback URL to your video player, and send updates whenever your product has a new idea.
worldstream.ts
const world = await sequencer.worldstreams.create({
prompt: 'A bioluminescent city beneath the ocean',
image_url: 'https://example.com/world.jpg',
});
// Add one web stream to your player
video.src = world.playback_url;
// Change what happens next
await world.update({
prompt: 'The city rises into the morning light',
});
API shape shown for early-access preview. Final endpoints and SDK syntax may change.
Built for continuous generation.
Worldstreams give interactive products a persistent video output and a direct control loop.
Image-to-video API
A generation job that returns a video clip
Each request creates a separate output
Your product waits for a completed file
New direction starts another generation job
Your application manages playback continuity
Worldstream API
A live session that keeps producing video
One session continues across many scenes
Viewers connect through one playback URL
Prompts and images update the active world
Generation and delivery run as one loop
From API request to living world.
A simple session model turns generative video into a responsive media surface.
01
Start a world
Send a prompt and an optional reference image to create a persistent generation session.
02
Play the stream
Use the playback URL in your product while Sequencer generates and delivers each new scene.
03
Update it live
Change the prompt, image, or creative direction through another API request.
04
Keep the world moving
The active session carries the new direction into the scenes that follow.
Give every product a video world.
Use a Worldstream anywhere software needs to generate, show, and redirect video in real time.
Interactive entertainment
Let an audience vote, prompt, and steer a shared video world.
Agent-generated channels
Give an AI agent a video output it can direct through API calls.
Live creative products
Build story apps, virtual environments, games, and installations around a living stream.
Build the next live video experience.
Tell us what you want to stream. Early-access partners can help shape the API, SDK, and delivery model.
Request early access
© 2026 Sequencer. All rights reserved.
Solutions
For Enterprises & Brands
Marketing Studio
Private Mode
For Filmmakers
Experiences
Voice Actor Pro
Live Avatar
Worldstreams
Worldstream API
Studio
Pricing
Community
Extensions
Model Arena
Tutorials
Resources
Contests
Film Festival
Webinars
Job Board
Affiliate Program