Video

Kling video (native format)

Generate Kling videos using Kling's own request format.

POST /kling/v1/videos/text2video

Use these endpoints if your code already speaks the official Kling API. The request body uses Kling's field names; RuToken forwards them to Kling and returns a task ID. For new integrations, the unified video API is simpler and works with every provider.

Method	Path	Purpose
`POST`	`/kling/v1/videos/text2video`	Text-to-video.
`POST`	`/kling/v1/videos/image2video`	Image-to-video.
`GET`	`/kling/v1/videos/text2video/{task_id}`	Query a text-to-video task.
`GET`	`/kling/v1/videos/image2video/{task_id}`	Query an image-to-video task.

Authentication

Send your RuToken key in the Authorization: Bearer sk-xxxx header. You do not need Kling access keys or JWT signing — the gateway handles that.

Request body

Content type: application/json. Fields not listed below are also passed through to Kling.

Parameter	Type	Required	Description
`model_name`	string	Yes	Kling model, e.g. `kling-v1`, `kling-v1-6`, `kling-v2-master`. `model` is accepted as an alias.
`prompt`	string	Yes	Positive prompt, up to 2500 characters. Required by the gateway for both text-to-video and image-to-video.
`negative_prompt`	string	No	What to avoid in the video.
`image`	string	Image-to-video	First frame as a URL or Base64 string (without the `data:` prefix). When present, the task is image-to-video.
`image_tail`	string	No	Last frame as a URL or Base64 string.
`mode`	string	No	`std` (default) or `pro`.
`duration`	string	No	`"5"` (default) or `"10"`. Must be a string.
`aspect_ratio`	string	No	`16:9`, `9:16` or `1:1`. Text-to-video only.
`cfg_scale`	number	No	Prompt adherence from `0` to `1`. Default `0.5`.
`camera_control`	object	No	Camera movement, as defined by Kling (`type`, `config`).
`static_mask` / `dynamic_masks`	string / array	No	Motion brush masks for image-to-video.
`external_task_id`	string	No	Your own task ID, stored by Kling.

Request example

cURL

curl https://rutoken.ai/kling/v1/videos/text2video \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $API_KEY" \
  -d '{
    "model_name": "kling-v1",
    "prompt": "A red fox running through a snowy forest, slow motion",
    "negative_prompt": "blurry",
    "mode": "std",
    "duration": "5",
    "aspect_ratio": "16:9"
  }'

Image-to-video:

cURL

curl https://rutoken.ai/kling/v1/videos/image2video \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $API_KEY" \
  -d '{
    "model_name": "kling-v1",
    "image": "https://example.com/portrait.jpg",
    "prompt": "The person smiles and waves at the camera",
    "duration": "5"
  }'

Response

json

{
  "id": "task_6Fh2kVb0mQ9xYzA1pL3sT8uWcR4nE7dJ",
  "task_id": "task_6Fh2kVb0mQ9xYzA1pL3sT8uWcR4nE7dJ",
  "object": "video",
  "model": "kling-v1",
  "status": "queued",
  "progress": 0,
  "created_at": 1760000000
}

Use task_id to query the task. It is a RuToken task ID, not the Kling task ID.

Query a task

cURL

curl https://rutoken.ai/kling/v1/videos/text2video/task_6Fh2kVb0mQ9xYzA1pL3sT8uWcR4nE7dJ \
  -H "Authorization: Bearer $API_KEY"

json

{
  "code": "success",
  "message": "",
  "data": {
    "task_id": "task_6Fh2kVb0mQ9xYzA1pL3sT8uWcR4nE7dJ",
    "action": "textGenerate",
    "status": "SUCCESS",
    "progress": "100%",
    "fail_reason": "",
    "result_url": "https://example.com/kling/output.mp4",
    "submit_time": 1760000000,
    "finish_time": 1760000150,
    "data": { "code": 0, "data": { "task_status": "succeed", "task_result": { "videos": [ { "url": "https://example.com/kling/output.mp4", "duration": "5" } ] } } }
  }
}

The response format and status values (SUBMITTED, QUEUED, IN_PROGRESS, SUCCESS, FAILURE) are the same as for GET /v1/video/generations/{task_id}; data.data contains Kling's original task response. The task can also be queried through GET /v1/video/generations/{task_id} or GET /v1/videos/{task_id}.