Kling 3.0

Kling 3.0 is a family of video models: realistic motion and scenes from text or a starting frame, with audio. Two tiers with identical parameters: Standard (kwaivgi/kling-v3.0-std) — a balance of price and quality, and Pro (kwaivgi/kling-v3.0-pro) — top-tier motion quality and detail. Generation is asynchronous: POST /v1/videos returns a job, then you poll the status and download the mp4.

Parameters

Field	Description
model	kwaivgi/kling-v3.0-std or kwaivgi/kling-v3.0-pro (the short kling-v3.0-std / kling-v3.0-pro; the alias kling = Pro)
prompt	Scene description, required
seconds	One of 5, 10, 15; defaults to 5
size	One of the three sizes below; defaults to 1280x720
input_reference	Image-to-video: one image as the starting frame, {"image_url": "data:image/...;base64,..."}
input_references	Not supported: the models have references set to false; a request with the array returns 400
seed	Not supported: the model’s seed is false; a request with the field returns 400

All three sizes are 720p and differ only in frame format:

Aspect	Size
16:9	1280x720
9:16	720x1280
1:1	720x720

Examples

Text-to-video

Python

import requests

job = requests.post(
    "https://api.mixen.ai/v1/videos",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "kwaivgi/kling-v3.0-std",
        "prompt": "A dancer in a bright bucket hat performs a breakdance on a concrete lot, a long shadow from the setting sun",
        "seconds": "10",
        "size": "1280x720",
    },
).json()
print(job["id"], job["status"])

cURL

curl https://api.mixen.ai/v1/videos \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kwaivgi/kling-v3.0-std",
    "prompt": "A dancer in a bright bucket hat performs a breakdance on a concrete lot, a long shadow from the setting sun",
    "seconds": "10",
    "size": "1280x720"
  }'

Image-to-video from a starting frame

The image goes in as a single data: URL — external links are not accepted:

import base64
import requests

with open("portrait.jpg", "rb") as f:
    img = "data:image/jpeg;base64," + base64.b64encode(f.read()).decode()

job = requests.post(
    "https://api.mixen.ai/v1/videos",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "kwaivgi/kling-v3.0-pro",
        "prompt": "The portrait comes alive: the gaze shifts aside, a faint smile, hair stirring",
        "seconds": "5",
        "size": "1280x720",
        "input_reference": {"image_url": img},
    },
).json()
print(job["id"], job["status"])

Polling and download

curl https://api.mixen.ai/v1/videos/video_... \
  -H "Authorization: Bearer $MIXEN_API_KEY"

curl -L https://api.mixen.ai/v1/videos/video_.../content \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -o kling.mp4

Statuses: queued → in_progress → completed or failed. A sensible polling interval is 5–10 seconds. Full cycle — video guide.

Price

Both tiers are billed per second, and the rate does not depend on the frame format: Standard — 16.8453 ₽ ($0.189) per second, Pro — 22.4604 ₽ ($0.252). You are charged duration × the tier’s rate. Current prices — catalog.

Strengths and limits

• Realistic motion and scenes from text or a frame — the family’s profile; the audio is already in the finished clip.
• Two tiers for the task: Standard for everyday scenes, Pro when maximum motion detail matters.
• Three frame formats: landscape, vertical, and square.
• 720p only: the family has no 480p or 1080p.
• Durations are fixed: only 5, 10, or 15 seconds — no continuous range like Seedance’s.
• Reference-to-video is not supported; an image can only be passed as the starting frame input_reference.

All models — in the catalog. The general video workflow — video guide.