How to Use the Seedance API — Pricing and Python Implementation (Seedance 2 Support, 2026)

Seedance is ByteDance’s text-to-video family, and the current-generation Seedance 2 is a notable 2026 model with native audio, multi-shot composition, and reference-image consistency. The Seedance API is available on Kunavo — an OpenAI-compatible endpoint, per-second USD billing, and no Volcano Engine account required.

Seedance API pricing

Billing is by generated video seconds, by resolution:

Billing is in USD per clip, deducted from prepaid Stripe balance — no credit-bundle conversion is needed. Rates are displayed from the same source as the pricing page.

How to Use — Generate Video from Text

Just send a prompt to POST /v1/video/generations. seedance-2-fast is recommended for testing, and seedance-2 for production assets:

import requests

resp = requests.post(
    "https://api.kunavo.com/v1/video/generations",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "seedance-2-fast",   # 最高品質は "seedance-2"
        "prompt": "折り鶴がほどけて鳥の群れになる、マクロ撮影、朝の柔らかい光",
        "duration": 5,
        "aspect_ratio": "16:9",
        "resolution": "720p",         # Fast: 480p/720p、Seedance 2 は 1080p も可
    },
    timeout=600,  # 生成には数分かかることがあります
)
print(resp.json()["data"][0]["url"])

Generation takes several minutes. In production, use the asynchronous /v1/videos task flow and include Idempotency-Key so retries do not cause duplicate charges.

Image to Video

Pass image_url and the still image becomes the first frame. Reference consistency is one of Seedance’s core strengths:

# 画像から動画へ: image_url を渡すと、その静止画が最初のフレームになります。
resp = requests.post(
    "https://api.kunavo.com/v1/video/generations",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "seedance-2",
        "prompt": "商品がゆっくり回転し、スタジオライトが表面を滑る",
        "image_url": "https://example.com/product-hero.png",
        "duration": 5,
        "resolution": "720p",
    },
    timeout=600,
)

Choosing Between Seedance 2 and Fast

Fast is the high-volume version of the family — low latency, about 20% cheaper, capped at 720p, and suited to social media and bulk-generation pipelines. Seedance 2 is for hero assets requiring 1080p and multi-shot consistency. The request format is identical, so switching requires changing one word in model.

Comparison with Veo 3 and Wan

All run on the same endpoint. Choose Seedance 2 for storytelling and reference consistency, Veo 3 for overall quality and native audio (Lite $0.18/clip~), and Wan 2.7 for the cheapest cinematic output ($0.096/second). See the text-to-video API guide for the full picture.

Frequently asked questions