{

}

[

]

video.dev

<video.dev/>
The Programmable Video Stack for theAI-Native World

vision.js

// The Vision

goal

We're building the programmable video layer for a world powered

technologies

by LLMs, agents, and multimodal AI

approach

Not just apps — tools, workflows, and systems built with and around video

// The future of video is code

vision.js

video.dev.ts

ai.config.js

// The Vision

import { Video } from "@video.dev/core";

const vision = {

goal: "We're building the programmable video layer for a world powered",

technologies: "by LLMs, agents, and multimodal AI",

approach: "Not just apps — tools, workflows, and systems built with and around",

focus: "video"

6};

7// The future of video is code

Shape the future of building with AI + video

The New Paradigm

Video Is the Most Powerful Output of AINow It's Finally Programmable

In the post-AI world, developers don't just generate text or code — they generate experiences. Our platform gives you the tools to treat video like code: editable, composable, and intelligent.

APIs

Generate, edit, and render video with simple API calls

POST /api/video/generate

Workflow Engine

Orchestrate AI agents, voice, avatars, and media seamlessly

workflow.connect(agents, media)

SDKs & Tools

Integrate into anything—from internal tools to full-blown SaaS

import { Video } from "@video.dev/sdk"

Video Understanding

Built-in indexing, retrieval, and semantic search capabilities

video.search("keyword")

Multimodal Support

Connect prompts, docs, voice, and visuals into coherent video

combine(text, image, audio)

Vector Embeddings

Turn video content into searchable, queryable data assets

video.getEmbeddings()

The building blocks are here — we're making them developer-first

Real-World Applications

What Developers Will Build(and Are Already Prototyping)

The next wave of AI-native applications is being built with video as a core interface.

Education

Why This Platform Exists

❝

❞

Because building with video has been stuck in 2015: slow editors, timeline UIs, and manual exports.Because AI changed what's possible — and devs need a new set of tools.Because code is better than clicking, and automation is better than repetition.

We're building a video infrastructure platform for:

LLM-Native Products

Applications built from the ground up with LLMs as a core component

AI/Agent Ecosystems

Platforms where multiple AI agents collaborate and interact

Multimodal Workflows

Systems that seamlessly blend text, image, audio, and video

API-First Teams

Developer teams who build with APIs and code first

You bring the ideas. We give you the building blocks.

The Future of Video Isn't Rendered. It's Coded.

This isn't just an API. It's a launchpad for the next wave of:

SaaS tools

Autonomous workflows

Media agents

Data-native interfaces

Dynamic content

Search systems

const future = await video.createFromCode({
  ecosystem: ai-native,
  programmable: true
});

Early Adopters Program

Want Early Access?

We're building this with developers, indie hackers, researchers, and ambitious founders.Join early and help shape the video infrastructure layer of the AI-native internet.

Developer Preview

Get early API access and influence our development roadmap.

Direct Support

Work directly with our engineering team as you build.

Founder Pricing

Lock in special pricing available only to early adopters.

constvideo=createVideo("Hello Video.dev!");

<video.dev/>The Programmable Video Stack for theAI-Native World

Video Is the Most Powerful Output of AINow It's Finally Programmable

APIs

Workflow Engine

SDKs & Tools

Video Understanding

Multimodal Support

Vector Embeddings

What Developers Will Build(and Are Already Prototyping)

AI Tutors with Adaptive Content

Searchable Video Libraries

Meeting Intelligence

Document-to-Video Pipelines

Programmable Story Agents

Video Workflow Automation

Personalized Marketing Video

Video Search Indexing

Multimodal Data Visualization