> ## Documentation Index
> Fetch the complete documentation index at: https://docs.anyfast.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# kling-v2-6

> Use Kling 2.6 text-to-video, image-to-video, motion control, and voice management through the Kling Legacy APIs.

Kling 2.6 is Kuaishou's video generation model, available through AnyFast with the Kling Legacy API structure. It supports text-to-video, image-to-video, native audio, motion control, and reusable voices.

## Key capabilities

* **Text to video** — Generate video from a prompt in `16:9`, `9:16`, or `1:1`.
* **Image to video** — Use a start frame, an end frame, or both, with motion brush and camera controls.
* **Native audio** — Set `sound` to `on` to generate synchronized audio.
* **Specified voices** — Reference up to two voices with `<<<voice_1>>>` and `<<<voice_2>>>` in the prompt.
* **Motion control** — Transfer motion from a reference video to a character image.
* **Standard and professional modes** — `std` produces 720p and `pro` produces 1080p.

<Note>
  Kling 2.6 supports fixed 5-second and 10-second durations. It does not support Kling 3.0 multi-shot generation, flexible 3–15 second duration, or 4K mode.
</Note>

## Text to video

```bash cURL theme={null}
curl --request POST \
  --url https://www.anyfast.ai/kling/v1/videos/text2video \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --header 'Content-Type: application/json' \
  --data '{
    "model_name": "kling-2.6",
    "prompt": "A corgi runs along a wet beach at sunset",
    "negative_prompt": "blurry, shaky",
    "duration": "5",
    "mode": "pro",
    "sound": "on",
    "aspect_ratio": "16:9"
  }'
```

## Image to video

```bash cURL theme={null}
curl --request POST \
  --url https://www.anyfast.ai/kling/v1/videos/image2video \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --header 'Content-Type: application/json' \
  --data '{
    "model_name": "kling-2.6",
    "image": "https://example.com/first-frame.png",
    "prompt": "The person smiles and waves at the camera",
    "duration": "5",
    "mode": "pro",
    "sound": "on"
  }'
```

Images can be public URLs or raw Base64 strings without a `data:image/...;base64,` prefix. Use JPG, JPEG, or PNG files no larger than 10 MB, with each side at least 300 px and an aspect ratio from 1:2.5 through 2.5:1.

## Motion control

Motion control combines a character image with a motion reference video. Reference videos must be MP4 or MOV, no larger than 100 MB, and at least 3 seconds long. The maximum is 10 seconds when `character_orientation` is `image` and 30 seconds when it is `video`.

<Warning>
  For difficult or fast motion, the model may extract only the usable continuous segment and return a result shorter than the uploaded video. A task can still run when at least 3 seconds of continuous usable motion is found, and the consumed credits are not refunded. Reduce the motion difficulty or speed when needed.
</Warning>

## Voice management

Create a reusable custom voice from a public MP3, WAV, MP4, or MOV URL, or from an eligible previously generated video. The source should contain one clear speaker and be 5–30 seconds long. Use the returned `voice_id` in a video's `voice_list`.

<Card title="Kling 2.6 API reference" icon="code" href="/api-reference/model-api/kuaishou/kling-v2-6">
  View the complete Legacy API parameters for text to video, image to video, motion control, and voice management.
</Card>

<script src="/feedback.js" />