QuickstartMarkdown

Get the result

GET /v1/generations/{id} returns the generation. Poll it every few seconds until status is succeeded or failed. Polling is free and never counts against a rate limit.

ID=gen_4Fq9xT2mLr7Kd1sV8bN3pZ0c

curl -s https://vidlab.ai/v1/generations/$ID \
  -H "Authorization: Bearer $VIDLAB_API_KEY"

import os, requests

job_id = "gen_4Fq9xT2mLr7Kd1sV8bN3pZ0c"
r = requests.get(
    f"https://vidlab.ai/v1/generations/{job_id}",
    headers={"Authorization": f"Bearer {os.environ['VIDLAB_API_KEY']}"},
)
job = r.json()["data"]
print(job["status"], [o["url"] for o in job["outputs"]])

const jobId = "gen_4Fq9xT2mLr7Kd1sV8bN3pZ0c";
const res = await fetch(`https://vidlab.ai/v1/generations/${jobId}`, {
  headers: { Authorization: `Bearer ${process.env.VIDLAB_API_KEY}` },
});
const { data: job } = await res.json();
console.log(job.status, job.outputs.map((o) => o.url));

A finished generation:

json

{
  "data": {
    "id": "gen_4Fq9xT2mLr7Kd1sV8bN3pZ0c",
    "object": "generation",
    "status": "succeeded",
    "model": "seedance-2.5",
    "variant": "720p",
    "units": 5,
    "price_usd": 1.05,
    "outputs": [{ "url": "https://...", "mime": "video/mp4" }],
    "created_at": "2026-11-01T18:04:11.000Z",
    "finished_at": "2026-11-01T18:06:02.000Z"
  }
}

Poll every 3 seconds for images and every 5 to 10 for video. VidLab asks the provider on its own schedule, so polling faster does not finish sooner.
GET /v1/jobs/{id}, /v1/jobs/{id}/status and /v1/jobs/{id}/result return the same answer.
Another account's id answers 404 generation_not_found.

Download the file

The files are in outputs. Download them with your key, so you never depend on a link that expires: GET /v1/generations/{id}/content is the first file, /content/2 the second, and so on. The answer redirects to the stored file, so follow redirects (curl -L; fetch and requests do by default).

curl -sL https://vidlab.ai/v1/generations/$ID/content \
  -H "Authorization: Bearer $VIDLAB_API_KEY" -o out.mp4

import os, requests

job_id = "gen_4Fq9xT2mLr7Kd1sV8bN3pZ0c"
file = requests.get(
    f"https://vidlab.ai/v1/generations/{job_id}/content",
    headers={"Authorization": f"Bearer {os.environ['VIDLAB_API_KEY']}"},
)
open("out.mp4", "wb").write(file.content)

import { writeFile } from "node:fs/promises";

const jobId = "gen_4Fq9xT2mLr7Kd1sV8bN3pZ0c";
const file = await fetch(`https://vidlab.ai/v1/generations/${jobId}/content`, {
  headers: { Authorization: `Bearer ${process.env.VIDLAB_API_KEY}` },
});
await writeFile("out.mp4", Buffer.from(await file.arrayBuffer()));

A generation that is not finished answers 409 not_ready (with Retry-After), a failed one 409 generation_failed, and a number past the last file 404 output_not_found.
The url in outputs is a signed link that lasts 6 hours; each poll returns a working one. The files themselves are kept for 30 days after the generation finished (outputs_kept_until). Copy files you want to keep longer to your own storage.
An image call with n above 1 returns one output per image.

List generations

GET /v1/generations returns your generations, newest first, with the same key. It is free and never rate limited. Use it for a task log, or to find an id you lost.

Query	Values	What it does
`limit`	1 to 100 (default 25)	How many generations in the page.
`starting_after`	A generation id	The next page: send the last id of the page before.
`status`	`queued`, `running`, `succeeded`, `failed`	Only generations in that status.
`model`	A model id	Only that model's generations.

curl -s "https://vidlab.ai/v1/generations?limit=10&status=succeeded" \
  -H "Authorization: Bearer $VIDLAB_API_KEY"

The answer is { "object": "list", "data": [generation, ...], "has_more": false }. When has_more is true, call again with starting_after set to the last id. A bad query value answers 400 (invalid_limit, invalid_status, invalid_model, invalid_starting_after).