Create API
AI Video Generator API
Combine AI video, images, voices and music with your existing media to produce thousands of videos programmatically. One AI video generator API for your entire pipeline.
Models and capabilities
-
Wan 3.0 Prime “Seen from above, morning sunlight slides across a green desk and steam curls up from a cup” AI Video Generation
Text-to-Video AI API
Describe a scene and the clip is generated during the render. Seedance 2.0 is the default; Seedance 2.5 makes clips up to 30 seconds long, with optional synchronised audio.
-
Wan 3.0 Prime, from a GPT Image 2.5 still “A breeze lifts the shirt’s hem and sleeves, which sway and slowly settle, as the camera pushes in” Image to Video
AI Image to Video API
Bring your images to life. Provide the URL of an image, and AI turns it into a video to use in your edits. Add an end image to guide the final frame.
-
GPT Image 2.5 Sunburst “A plain paper shopping bag on a mustard-yellow backdrop, in hard morning light” AI Image Generation
Text-to-Image AI API
Generate images from a text prompt. Nano Banana 2 is the default, and one field switches models. Edit models change an existing image from a written instruction.
-
ElevenLabs Multilingual v2, Matilda “Welcome home. Here’s your first look at twelve Harbour Street.” AI Voiceovers
Text-to-Speech AI API
Generate voice-overs from your script with ElevenLabs, MiniMax or Amazon Polly voices. ElevenLabs Multilingual v2 speaks 29 languages.
-
ElevenLabs Music, instrumental “Low pulsing strings over a steady kick at 120 BPM, building to a big final swell” AI Music
AI Music Generation API
Describe the mood, tempo and instruments, with or without vocals, and ElevenLabs Music generates a track for your edit.
One request, not a pipeline
Putting AI-generated assets in a video means plumbing, and every provider hands it back to you: poll the job, download the file before its link expires, host it, pass the URL to your video API. Multiply that by every model you use.
In Shotstack you put the prompt in the render request. Nothing else to wire up.
Per provider, every time: 7 steps
- Open an account and get an API key
- Create a generation job
- Poll until it finishes, fails or stalls
- Download the output before the link expires
- Host the file
- Pass its URL to your video API
- Render
With Shotstack: 2 steps
-
Add a prompt to the asset
{ "type": "image", "model": "gpt-image-2.5-sunburst", "prompt": "A matte speckled stoneware mug…" } - Render
Automate via API
One AI video API
Text-to-video, image-to-video, text-to-image, speech and music run through the same video generation API as the edit. Add a prompt to an asset, choose the model, and the render generates it, cuts it in and delivers the finished file.
Unchanged prompts are reused at no charge. Generated files stay in your account, on our CDN, until you delete them.
Generated in one render
- Text-to-image
gpt-image-2.5-sunburstA woven rattan basket holding a rolled linen blanket, on a mustard-yellow seamless paper backdrop, in warm daylight. …
- Image-to-video
wan-3.0-prime-image-to-videoWarm daylight brightens slowly across the woven rattan. The camera holds still and the framing stays the same.
- Speech
elevenlabs-multilingual-v2Our rattan basket is back. Woven by hand, and useful everywhere.
- Music
elevenlabs-musicWarm, unhurried acoustic guitar with soft brushed percussion at 90 BPM. Light, modern and clean, for short product videos.
Result
Built for volume
AI video creation API for scale
One template, one row of data per video. Thousands render concurrently, a webhook fires when each is done, and the file lands on our CDN or in your own storage.
The same engine runs your first test and your ten-thousandth video.
Trusted to deliver
Join companies large and small rendering thousands of videos every day
Over 100,000 Businesses from 119 Countries Trust Shotstack
Achieving the desired outcome would have required extensive research and at least two to six months of engineering time.
The Shotstack AI video generation API
One render request generates the clips, images, voice and music and edits them into the finished video.
Questions
What is an AI video generation API?
An API that creates video, images, voice or music from a prompt and returns a file. In Shotstack it is part of the render: put a prompt and a model on any asset in the timeline, and the asset is generated during the render and edited into the finished video. One AI video generator API, one request.
Is there a free AI video generator API tier?
New accounts get 10 free credits, valid for 30 days, with no credit card. Generation is charged per model from the same credit balance as rendering, so the free credits cover your first generated clips, images and voice-overs.
How is AI generation priced?
Each generation costs what the provider of the model charges plus a small margin, converted to credits. Video is charged per second, speech per character and music per started minute, at the same rate on every plan. See all models and pricing.
Can I see the price before I generate?
Yes. Send the asset to the quote endpoint and it returns the cost in credits. Nothing is generated, reserved or charged. An asset identical to one you generated before costs nothing.
Which models can I use?
Video from Seedance, Wan and Gemini, images from Nano Banana, GPT Image and FLUX, voices from ElevenLabs, MiniMax and Amazon Polly, and music from ElevenLabs Music. One field on the asset switches the model. See all models and pricing.
Can I start from my own image?
Yes. Image-to-video models take a start image by URL and an optional end image to guide the final frame, so a product photo or a listing shot becomes the clip.
What happens if a generation fails?
The render fails with the reason and that generation is not charged. Assets that did generate are kept, so a retry only makes what is missing, and an unchanged prompt reuses its asset at no charge.
How do I get an API key?
Sign up and your keys are in the dashboard: one for the sandbox, one for production. Request API keys.
Can I test in the sandbox?
Yes. Sandbox renders are free and watermarked. Generation in the sandbox is charged like production, from your production credit balance, and the two environments store generations separately, so an asset made in the sandbox is charged again the first time production generates it.
More to do
-
AI video creation platform
Video workflows and applications built on the API, templates and AI generation.
-
Video editing API
Render video from JSON: templates, merge fields and the same request that generates AI assets.
-
Online video editor
Design the templates your renders fill, in the browser.
Your first render in minutes
- Free AI video generator API tier: 10 credits
- No credit card required