Create video · DeutschlandGPT

Starts an asynchronous text-to-video generation job. Available models include EU-hosted Google Veo (on Vertex AI) and fal.ai models (kling-2.6-pro, seedance-2.0, seedance-2.0-fast, seedance-1-pro, hailuo-02-pro) — swap the model string to change provider. Use /v2/models/all to list every video model.

Returns a job descriptor with status: "pending". Poll GET /v2/videos/generations/{id} until status is succeeded (returns signed URLs) or failed (returns an error message). Generation typically takes 30 seconds to several minutes.

Credits are charged up-front based on the requested resolution, duration, and number of videos. If the job fails or produces fewer videos than requested, the difference is refunded automatically. fal.ai models accept extra knobs (negative_prompt, generate_audio, cfg_scale, …) via the parameters passthrough.

POSThttps://api.deutschlandgpt.de/v2/videos/generationsTry it

Example request

curl https://api.deutschlandgpt.de/v2/videos/generations \
  -X POST \
  -H "Authorization: Bearer $DGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "prompt": "A golden retriever puppy bounding through autumn leaves, slow motion, cinematic",
  "model": "veo-3.1-fast",
  "n": 1,
  "durationSeconds": 8,
  "resolution": "720p",
  "aspectRatio": "16:9",
  "seed": 0,
  "parameters": {}
}'

curl https://api.deutschlandgpt.de/v2/videos/generations \
  -X POST \
  -H "Authorization: Bearer $DGPT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "prompt": "A golden retriever puppy bounding through autumn leaves, slow motion, cinematic",
  "model": "veo-3.1-fast",
  "n": 1,
  "durationSeconds": 8,
  "resolution": "720p",
  "aspectRatio": "16:9",
  "seed": 0,
  "parameters": {}
}'

import os, requests

response = requests.post(
    "https://api.deutschlandgpt.de/v2/videos/generations",
    headers={"Authorization": f"Bearer {os.environ['DGPT_API_KEY']}"},
    json={
        "prompt": "A golden retriever puppy bounding through autumn leaves, slow motion, cinematic",
        "model": "veo-3.1-fast",
        "n": 1,
        "durationSeconds": 8,
        "resolution": "720p",
        "aspectRatio": "16:9",
        "seed": 0,
        "parameters": {}
    },
)

print(response.json())

import os, requests

response = requests.post(
    "https://api.deutschlandgpt.de/v2/videos/generations",
    headers={"Authorization": f"Bearer {os.environ['DGPT_API_KEY']}"},
    json={
        "prompt": "A golden retriever puppy bounding through autumn leaves, slow motion, cinematic",
        "model": "veo-3.1-fast",
        "n": 1,
        "durationSeconds": 8,
        "resolution": "720p",
        "aspectRatio": "16:9",
        "seed": 0,
        "parameters": {}
    },
)

print(response.json())

const response = await fetch('https://api.deutschlandgpt.de/v2/videos/generations', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.DGPT_API_KEY}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    "prompt": "A golden retriever puppy bounding through autumn leaves, slow motion, cinematic",
    "model": "veo-3.1-fast",
    "n": 1,
    "durationSeconds": 8,
    "resolution": "720p",
    "aspectRatio": "16:9",
    "seed": 0,
    "parameters": {}
  }),
});

console.log(await response.json());

const response = await fetch('https://api.deutschlandgpt.de/v2/videos/generations', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.DGPT_API_KEY}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    "prompt": "A golden retriever puppy bounding through autumn leaves, slow motion, cinematic",
    "model": "veo-3.1-fast",
    "n": 1,
    "durationSeconds": 8,
    "resolution": "720p",
    "aspectRatio": "16:9",
    "seed": 0,
    "parameters": {}
  }),
});

console.log(await response.json());

Request body

promptstringrequired

Text description of the desired video. English produces the most reliable results.

modelstringoptional

Video model to use. Defaults to veo-3.1-fast. Use /v2/models to discover available video models.

nintegeroptional

Number of videos to generate. Capped per-model; Veo currently allows up to 4.

durationSecondsintegeroptional

Length of the video in seconds. Must be one of the model's supported durations (Veo: 4, 6, or 8).

resolutionstringoptional

Output resolution. Must be supported by the model (Veo: 720p, 1080p).

aspectRatiostringoptional

Aspect ratio. Must be supported by the model (Veo: 16:9, 9:16).

seedintegeroptional

Optional seed for reproducible generation (when supported by the model).

parametersobjectoptional

Provider-specific passthrough merged into the upstream request (fal.ai only). e.g. { "negative_prompt": "blurry", "generate_audio": true }. Ignored by Vertex/Veo models.

Response

200application/json

{
  "id": "00000000-0000-0000-0000-000000000000",
  "object": "video.generation_job",
  "status": "pending",
  "model": "string",
  "created": 0,
  "params": {
    "durationSeconds": 0,
    "resolution": "string",
    "aspectRatio": "string",
    "n": 0
  },
  "estimated_cost_in_euro": 0
}

{
  "id": "00000000-0000-0000-0000-000000000000",
  "object": "video.generation_job",
  "status": "pending",
  "model": "string",
  "created": 0,
  "params": {
    "durationSeconds": 0,
    "resolution": "string",
    "aspectRatio": "string",
    "n": 0
  },
  "estimated_cost_in_euro": 0
}

Job accepted; poll the GET endpoint with the returned id.

idstring<uuid>

objectstring

statusstring

modelstring

createdinteger

Unix timestamp

paramsobject

durationSecondsinteger

resolutionstring

aspectRatiostring

ninteger

estimated_cost_in_euronumber

Upper bound on cost charged at job creation. Refunded down to actual cost on completion.

Response codes

200

Job accepted; poll the GET endpoint with the returned id.

object

400

Invalid duration, resolution, or aspect ratio for this model

401

Unauthorized

402

Insufficient credits

403

API key lacks permission for this model

404

Model not found