MAKEUGC PLATFORM API · DOCS V2

Kling 4.0 · Text to video

Kling 4.0 · text-to-video

VIDEO OUTPUT

Request

POST /enterprise/generate

Model ID: kling-v4-t2v. Put model inputs inside params; keep model and project_id at the top level. Send webhook_url at the top level when using notifications.

Alias: kling/v4/text-to-video

Unsupported fields are rejected, including seed, cfg_scale, negative_prompt, generate_audio and client prices. Do not send mode, user_id, project_id or model_version in params.

Media accepts public HTTP(S) URLs or matching base64 data URIs. Images: up to 30 MiB, at least 300px on each side. Videos: up to 300 MiB, 700–4553px on each side, 23.9–60.05 FPS. Source/reference videos: 3–30.05 seconds; element videos: 3–8.05 seconds including metadata tolerance. Media aspect ratio must be 0.4–2.5. Actual bytes and video durations are checked before charging.

Generation without video input costs $0.157/s at 720p or $0.235/s at 1080p. Edit and reference generation with video use $0.094/s or $0.141/s and bill max(input seconds + output seconds, 2 × output seconds). Credits = ceil to 0.1(provider USD × 1.5 / 2.50). Current pricing_version: kling-v4-2026-10-27-v2.

Use the server quote: 5s generation is 0.5 credits at 720p or 0.8 at 1080p. Do not halve an old rounded quote. Pricing fields are response-only; never send retail_multiplier, pricing_version or credits in params. Enqueue and status include the saved pricing snapshot; historical jobs retain their original policy.

The initial quote is reserved once. Status returns credits_deducted, credits_charged and credit_adjustment/refund_result. Shorter outputs may receive a partial refund; final charges never exceed the reservation. Failed jobs refund the reservation.

Supply prompt or multi_prompt (or both). Each shot has a 1–512 character prompt and integer duration 1–30 (default 5). Total shot duration must be 3–30 seconds and overrides duration.

Request envelope

Parameter	Description
`model` string Required	Exact public model ID from the model catalog.
`params` object Required	Model-specific input fields. See the individual model page for required fields, types, defaults and limits.
`project_id` string Optional	Existing project belonging to your account. Omit to create a project automatically.
`webhook_url` string Optional	Optional callback URL for supported models. Models marked polling-only reject this field. Enterprise Seedance does not forward it. Always keep polling available.

Model parameters

Parameter	Description
`prompt` string Optional	Non-blank prompt; required unless multi_prompt is provided for generation. Allowed: `1–26000 Unicode characters; preserved without truncation`
`resolution` string Optional	Output resolution. Allowed: `720p \| 1080p` Default: `1080p`
`duration` integer Optional	Output seconds; overridden by the sum of multi_prompt durations. Allowed: `3–30` Default: `5`
`multi_prompt` object[] Optional	Non-empty ordered shots: {prompt, duration}. Total duration 3–30 seconds; no six-shot limit.
`aspect_ratio` string Optional	Output aspect ratio. Allowed: `16:9 \| 9:16 \| 1:1 \| 21:9` Default: `16:9`

Code examples

Set MAKEUGC_API_KEY on your server. Replace example media URLs with files you host and keep accessible during processing.

cURL

curl --fail-with-body --max-time 120 -X POST "https://app.makeugc.ai/api/platform/enterprise/generate" \
  -H "X-Api-Key: $MAKEUGC_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{
  "model": "kling-v4-t2v",
  "params": {
    "prompt": "A slow camera move around a ceramic coffee cup in warm morning light.",
    "resolution": "1080p",
    "duration": 5
  }
}'

JavaScript

// Run on your server with Node.js 20+.
const apiKey = process.env.MAKEUGC_API_KEY;
if (!apiKey) throw new Error("Set MAKEUGC_API_KEY first");

const response = await fetch("https://app.makeugc.ai/api/platform/enterprise/generate", {
  method: "POST",
  headers: {
    "X-Api-Key": apiKey,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
  "model": "kling-v4-t2v",
  "params": {
    "prompt": "A slow camera move around a ceramic coffee cup in warm morning light.",
    "resolution": "1080p",
    "duration": 5
  }
}),
  signal: AbortSignal.timeout(120000),
});
const text = await response.text();
if (!response.ok) {
  throw new Error(`HTTP ${response.status}: ${text} (request ${response.headers.get("x-request-id")})`);
}
const result = JSON.parse(text);
console.log(result);

Python

# Install the requests package before running.
import os
import requests

payload = {
    "model": "kling-v4-t2v",
    "params": {
        "prompt": "A slow camera move around a ceramic coffee cup in warm morning light.",
        "resolution": "1080p",
        "duration": 5
    }
}

response = requests.post(
    "https://app.makeugc.ai/api/platform/enterprise/generate",
    headers={"X-Api-Key": os.environ["MAKEUGC_API_KEY"]},
    json=payload,
    timeout=120,
)
if not response.ok:
    raise RuntimeError(f"HTTP {response.status_code}: {response.text} "
                       f"(request {response.headers.get('x-request-id')})")
print(response.json())

200 · Job accepted

Generation response

{
  "request_id": "request_123",
  "project_id": "project_123",
  "media": "video",
  "model": "kling-v4-t2v",
  "status": "queued",
  "message": "Enterprise API job queued"
}

Retrieve the result

Save request_id, then poll every 5–10 seconds. A successful generation request does not mean the output is ready.

cURL

curl --fail-with-body --max-time 120 -X GET "https://app.makeugc.ai/api/platform/enterprise/status/request_123?media=video" \
  -H "X-Api-Key: $MAKEUGC_API_KEY"

JavaScript

// Run on your server with Node.js 20+.
const apiKey = process.env.MAKEUGC_API_KEY;
if (!apiKey) throw new Error("Set MAKEUGC_API_KEY first");

const response = await fetch("https://app.makeugc.ai/api/platform/enterprise/status/request_123?media=video", {
  method: "GET",
  headers: {
    "X-Api-Key": apiKey,
  },
  signal: AbortSignal.timeout(120000),
});
const text = await response.text();
if (!response.ok) {
  throw new Error(`HTTP ${response.status}: ${text} (request ${response.headers.get("x-request-id")})`);
}
const result = JSON.parse(text);
console.log(result);

Python

# Install the requests package before running.
import os
import requests

response = requests.get(
    "https://app.makeugc.ai/api/platform/enterprise/status/request_123?media=video",
    headers={"X-Api-Key": os.environ["MAKEUGC_API_KEY"]},
    timeout=120,
)
if not response.ok:
    raise RuntimeError(f"HTTP {response.status_code}: {response.text} "
                       f"(request {response.headers.get('x-request-id')})")
print(response.json())

200 · Completed

Status response

{
  "request_id": "request_123",
  "media": "video",
  "status": 1,
  "video_url": "https://your-cdn.example/result.mp4",
  "created_at": "2026-10-13T00:00:00.000Z",
  "updated_at": "2026-10-13T00:00:30.000Z"
}

Read status: 0 is processing, 1 is completed and 2 is failed. Video status can also be a string. Output is in video_url . If completed has no output URL yet, continue polling within your timeout. See jobs & webhooks.

Credits & errors

Cost: quoted by resolution, output duration and measured video input. Read returned credit fields when available. See credit units and limits.

Invalid inputs return 400; insufficient credits can return 400 or 402 depending on the generation path. Access restrictions return 403 and temporary model unavailability can return 503. See error handling. Each POST can start a billable job; do not automatically resubmit after a timeout.

Related models

Browse all models →