NewLTX-2.3 is live — video with synchronized audio

Generate video at a fraction of the cost.

A webhook-first API for asynchronous AI video generation. We run your jobs on spot H100 GPUs — up to 70% cheaper than on-demand — then hand back an S3 link and a callback. You never touch a GPU.

No subscription · Pay per second of generated video · Only billed on success

Up to 70%
cheaper than on-demand
H100
GPUs, provisioned on spot
Audio
synchronized, with LTX-2.3
Webhooks
on every completed job

Two frontier models, one API

Pick the model that fits the job. Same submit flow, same webhooks, same per-second billing.

Wan 2.2

Image-to-video

High-quality video generation model from Alibaba.

Up to720p
Length15s
AudioNo
480p720p16:99:161:1
from$0.030 / sec

LTX 2.3

Image-to-video

Audio

Lightricks LTX-2.3 (22B) image-to-video with synchronized audio. Up to 20s clips at 24fps.

Up to1080p
Length20s
AudioYes
720p1080p16:99:161:1
from$0.016 / sec

Built for batch, priced for scale

We compete on cost and reliability — not on real-time latency. Perfect for pipelines that generate video in the background.

Spot cost arbitrage

We manage the full spot-instance lifecycle — provisioning, preemption recovery, and teardown — so you get on-demand reliability at spot prices. Flat per-second pricing, decoupled from the market.

Webhook-first async DX

Submit a job and move on. We POST to your webhook the moment the render lands in S3. No polling, no held connections — a clean primitive for n8n, Zapier, and custom backends.

Reliable by design

Durable SQS queues, automatic retries, dead-letter safety nets, and idempotent provisioning. Jobs survive spot preemption and are never silently dropped.

How it works

From image to finished clip in three steps.

01

Submit a job

Upload an image and a prompt via the dashboard or the API. Choose your model, resolution, duration, and quality tier.

02

We run it on H100s

Your job lands in a durable queue and we spin up a spot H100 to process it — then tear it down. You pay only for the seconds you generate.

03

Get your result

The finished video is uploaded to S3 and we fire a webhook to your endpoint. Download it from the dashboard or straight from the callback.

submit a job
POST https://getmecompute.com/api/jobs
{
  "model": "ltx-2-3",
  "prompt": "A calm ocean at golden hour, gentle waves",
  "image_url": "https://your-bucket.s3.amazonaws.com/frame.jpg",
  "resolution": "1080p",
  "duration": "10s",
  "quality": "standard",
  "webhook_url": "https://your-app.com/hooks/video-done"
}

Simple, transparent pricing

Priced per second of generated video and billed only on success. No subscriptions, no minimums.

Wan 2.2

ResolutionFasterStandardHighHighest
480p$0.030$0.040$0.050$0.060
720p$0.050$0.060$0.070$0.080

LTX 2.3

with audio
ResolutionFasterStandardHighHighest
720p$0.016$0.020$0.024$0.028
1080p$0.030$0.036$0.042$0.048

All prices are per second of generated video. Example: a 5-second Standard clip at $0.040/sec costs $0.20. Higher tiers use more denoising steps for richer motion & detail.

Ship video generation this afternoon.

Create an account, submit your first job, and get a webhook back. No sales call, no minimum spend.

Get Started