Seedance API — ByteDance's Seedance 2 text-to-video, live with pricing (2026)

Seedance is ByteDance's text-to-video family, and Seedance 2 is its current generation — native audio, multi-shot storytelling, and reference-image consistency that made it one of the most talked-about video models of 2026. The Seedance API generates short clips from a prompt or animates a still image, and it is live on Kunavo: OpenAI-compatible endpoint, per-second USD pricing about 15–20% under ByteDance's list, no Volcano Engine account.

Seedance API pricing

Kunavo serves two Seedance models, billed per second of generated video by resolution:

Billing is per clip in USD on a prepaid Stripe balance — no ByteDance credit bundles, no Volcano Engine console. Prices track the catalog, which is the same source the API bills from.

Generate a video with the Seedance API

Submit a prompt to POST /v1/video/generations. Start with seedance-2-fast for iteration, switch to seedance-2 for finals:

import requests

resp = requests.post(
    "https://api.kunavo.com/v1/video/generations",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "seedance-2-fast",       # or "seedance-2" for max quality
        "prompt": "a paper crane unfolding into a flock of birds, macro, "
                  "soft morning light, gentle camera pull-back",
        "duration": 5,
        "aspect_ratio": "16:9",
        "resolution": "720p",             # 480p / 720p (Fast), +1080p (Seedance 2)
    },
    timeout=600,
)
print(resp.json()["data"][0]["url"])

Generation takes a few minutes. In production, use the async /v1/videos task lifecycle — submit, store the task id, poll or receive the webhook — with idempotency keys so retries never double-bill.

Image-to-video with a first frame

Seedance's reference strength makes it a strong image-to-video model: pass image_url and the still becomes the clip's first frame:

# Image-to-video: animate a still by passing image_url —
# it becomes the first frame of the generated clip.
resp = requests.post(
    "https://api.kunavo.com/v1/video/generations",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "seedance-2",
        "prompt": "the product rotates slowly, studio light sweeps across",
        "image_url": "https://example.com/product-hero.png",
        "duration": 5,
        "resolution": "720p",
    },
    timeout=600,
)

Seedance 2 vs Seedance 2 Fast

Fast is the workhorse: same family, lower latency, ~20% cheaper per second, capped at 720p — right for social formats, previews and high-volume pipelines. Seedance 2 adds 1080p and the strongest multi-shot coherence — right for hero assets and narrative sequences. The request shape is identical; A/B testing is a one-word change.

Seedance vs Veo 3 vs Wan

All three are live on the same endpoint. Seedance 2 leads multi-shot narrative and character/reference consistency. Veo 3 leads overall cinematic quality and generates native synchronized audio (from $0.16/clip on Lite). Wan 2.7 is the budget cinematic pick at $0.096/second. For the full landscape, see the text-to-video API guide.

FAQ

What is the Seedance API?

ByteDance's text-to-video family, served on Kunavo as seedance-2 and seedance-2-fast via /v1/video/generations.

How much does it cost?

From $0.093/second (Fast 480p) to $0.612/second (Seedance 2 1080p) — about 15–20% under ByteDance's list; a 5s 720p clip is $0.99–1.23.

Does Seedance support image-to-video?

Yes — pass image_url and the still becomes the first frame; reference consistency is one of Seedance's core strengths.

Do I need a Volcano Engine account?

No. One Kunavo key covers Seedance plus Veo 3, Wan, image, music and text models, with per-clip USD billing.