What is Pipevideo?
Pipevideo is an OpenRouter-style API for AI video generation. Integrate once, then route requests to Kling, Runway, Minimax, and every major video model through a single, consistent endpoint.
Automatic fallbacks, provider-aware load balancing, and a unified SDK mean your code never changes when you swap models, add new providers, or scale across regions.
Ship AI video with one API
1. Login, add credits, and get an API key
Sign up or log in to your dashboard, add funds in billing, then create or copy your API key to authenticate every request.
Dashboard
Billing
Docs
2. Make your first API request
Call the generations endpoint with a model and prompt. Use the REST API or an SDK; we’ll return a job ID and then the video when it’s ready.
REST API
Models
3. Deliver your video
Get the finished video via webhook, polling, or real-time streams. Pipevideo returns a URL when the job completes.
Webhooks
Polling & streams
Why teams switch.
Teams switch to Pipevideo when traffic spikes are unpredictable and reliability matters more than owning GPU infrastructure.
You get model and provider fallback routing plus pay-per-generation pricing, so you avoid idle capacity and keep requests flowing.
"Pipevideo let us swap video providers without touching our application code. Fallbacks and routing just work, giving our team confidence that every video request lands successfully.ˮ
— Chris Bell, CTO, Knock
Video Generator Agent
Install and use Pipevideo with automatic fallbacks.
import { Pipevideo } from 'pipevideo';
import { NextResponse } from 'next/server';
const client = new Pipevideo({
apiKey: process.env.PIPEVIDEO_API_KEY
});
export async function POST(req: Request) {
const { prompt } = await req.json();
const video = await client.generations.create({
models: ['kling/3.0-pro', 'google/veo-3.1', 'minimax/video-01'],
prompt,
provider: {
order: ['fal', 'replicate'],
maxPrice: {perRequestCents: 500},
},
});
return NextResponse.json(video);
}