VEED API: Frequently Asked Questions

The VEED API gives you programmatic access to our AI video models. Send a request, get a video back. This article answers the most common questions about getting set up, how billing works, how jobs run, and what to do when something fails.

Each model does one job and is available on its own endpoint:

See all models at api.veed.io.

You can also test any model in the playground on its model page before writing any code. SDKs are available for Node and Go.

No. VEED API accounts and VEED editor accounts are separate. You only need an API account.

Manage your keys in the API keys section. Send your key in the Authorization: Bearer header on every request.

If a key is compromised, revoke it and create a new one.

The API is for business and professional use, including individual developers. You must be 18 or over.

Use the API if you're a developer building video features into your own product, or automating video at scale.

Use the VEED editor if you want to create or edit videos yourself, without writing code.

Each model has its own rate, charged in USD per second, per minute or per frame. See Pricing, or the Pricing tab on each model page — every model page also has a cost estimator.

The API runs on a credit system. Credits are prepaid, non-refundable, and expire two years after purchase. Prices exclude taxes.

No. API usage is billed separately, from your API credits. Your VEED subscription doesn't cover it.

In your API Dashboard: Activity shows individual jobs, and Billing shows your spend and remaining credits.

Yes. Contact sales for higher concurrency limits, direct access to our team, invoicing, and volume discounts.

Jobs are asynchronous. A request returns a job_id immediately. Poll the job's GET endpoint until the status is COMPLETED or FAILED.

Webhooks are not supported.

Yes. Limits apply per class of request: reads, writes, and AI generation jobs.

If you hit a limit you get a 429 error, and the response headers tell you when you can retry. For higher limits, contact sales.

Output URLs expire 24 hours after the job completes — not 24 hours after you first check it.

You can change this per request with the X-Veed-Media-Expiration-Seconds header, up to 30 days (2,592,000 seconds). Values above 30 days are capped rather than rejected.

Once a URL expires it returns 410 Gone and can't be renewed. Checking the job again returns the same expired link, so save anything you need to your own storage before it expires.

Tip: use a shorter expiry if the link goes somewhere public, such as an email or a log. Anyone with the link can download the file until it expires.

We store the request and response of each API call for 7 days so you can see them in your request log. After that, they're deleted.

To skip storing them for a specific call, send the header X-Veed-Store-IO: 0. This doesn't affect your output URLs, which follow the expiry rules above. We always keep basic details — time, model, status, duration and cost — for usage and billing.

We don't use your API inputs or outputs to train our models unless you agree to it.

Yes, you can use outputs commercially. See our Terms of Sale for further detail.

See our Code of Conduct. Only use images, videos and voices you own or have permission to use.

Jobs that break the rules fail with the content_moderation error.

A failed job returns an error code and message. These are the most common:

Full list in the error schema.

Email [email protected] or contact us via the support bubble in the bottom right corner. Include your job_id and the X-Request-ID from the response — it helps us find your request quickly.