Does Seedance do image-to-video as well as text-to-video?

Yes. Every Seedance tier on this page accepts a seed image to animate, as well as a text prompt.

How long can a Seedance clip be, and at what resolution?

Each tier's row states its duration ceiling and resolution; the newer line runs longer. Output defaults to the resolution on this page with synchronized audio on the text-to-video path.

How is Seedance priced through BlockRun?

Per second of generated video at the tier's rate — each tier's row shows the rate and the price of a default-length clip — quoted in the 402 before generation starts.

Can I keep the same character across Seedance clips?

Yes. Enroll a character once, as a Virtual Portrait or through RealFace, then pass its asset id on each request; the tiers that support it are marked on this page.

Does Seedance output include audio?

Text-to-video output ships with synchronized audio by default, matching what ByteDance's own product returns. Image-to-video follows the model's own behaviour, stated on its row.

Seedance via BlockRun vs Runway, Luma or fal.ai — which should I use?

Runway and Luma sell their own models on a subscription; fal.ai hosts others' behind a key and a balance. Here Seedance's tiers are paid per clip with no account, beside Sora and Grok Imagine on the same endpoint.

How do I use the Seedance video API without a BytePlus account?

Call BlockRun's video endpoint with a Seedance model. It generates the clip on ByteDance's Seedance family, priced per clip and paid by the request — no BytePlus or JiMeng account and no API key.

Can an AI video pipeline call Seedance unattended?

Yes. The request pays for itself and returns a job to poll; the finished MP4 lands on a hosted URL that does not expire, so the next step in the pipeline can take it directly.

All Services By capability: AI Video Generator

PARTNER

ByteDance Seedance

AI Video Generation for Agents

Seedance is ByteDance's video generation model family. Generate MP4 clips from text or a seed image — up to 15 seconds on the 2.0 line, 30 on 2.5 — at 720p with synchronized audio. Lock a character across clips with one-time enrollment. Priced per clip and quoted before it runs — no BytePlus account, no API keys.

5 model tiers live on Base

720p + synced audio

Website RealFace Guide

Capabilities

Text-to-Video

Generate MP4 clips from a text prompt — up to 15s on the 2.0 line and up to 30s on Seedance 2.5. Defaults to 720p, 5 seconds, with synchronized audio on the t2v path.

Image-to-Video

Pass a seed image to animate it. All five Seedance tiers accept an image input for image-to-video generation.

Character Consistency

Enroll a character once (Virtual Portrait or RealFace), then pass its ta_xxx asset id to keep the same character across every clip.

Synced Audio

Text-to-video output ships with synchronized audio by default — matching what you get from BytePlus / JiMeng directly.

Model Tiers

Five tiers, token-priced per second. No subscriptions — you pay only for the clips you generate. A 5-second clip is the default.

Seedance 1.5 Pro

bytedance/seedance-1.5-pro

~$0.35 / 5s clip

$0.070/s

Budget tier. Image input supported. Does NOT support RealFace assets.

No RealFace

Seedance 2.0 Fast

bytedance/seedance-2.0-fast

~$0.83 / 5s clip

$0.165/s

Fast generation (~60–80s). Image input + RealFace assets supported.

RealFace ✓

Seedance 2.0 Mini

bytedance/seedance-2.0-mini

~$0.40 / 5s clip

$0.080/s

Half the flagship rate. 480p / 720p only, up to 15s. Image input + RealFace assets supported.

RealFace ✓

Seedance 2.0 Pro

bytedance/seedance-2.0

~$1.14 / 5s clip

$0.227/s

Premium quality, up to 4K. Image input + RealFace assets supported.

RealFace ✓

Seedance 2.5

bytedance/seedance-2.5

~$1.58 / 5s clip

$0.315/s

Long-form: up to 30s, 720p ceiling. Image input supported. Does NOT support RealFace assets or reference media.

No RealFace

Keep the Same Character

Enroll a character once and reuse its ta_xxx asset id on every Seedance 2.0 / 2.0-fast / 2.0-mini call to lock the character across clips. Neither flow asks for a government ID or an account login. Putting a real person in AI video walks the liveness flow end to end, including the step that needs their phone.

Virtual Portrait

Upload an AI-generated character image, get back a ta_xxx asset id. For synthetic characters — no real-person verification needed.

$0.011 / enrollment (one-time)

RealFace

For a real person whose likeness you may use. The rights-holder completes a 60-second on-phone liveness check (nod + blink). Biometric data never touches BlockRun.

$0.011 / enrollment (one-time)

API Reference

Base (USDC): https://blockrun.ai/api/v1/videos/generations

Async submit → poll. POST a prompt + model to create a job; the response id (e.g. vd_abc123) is polled until the MP4 is ready. Send without a payment header to get a 402 with the exact price.

# 1. Submit (returns 402 first; sign EIP-3009 USDC TransferWithAuthorization)
curl -X POST https://blockrun.ai/api/v1/videos/generations \
  -H "Content-Type: application/json" \
  -H "x-payment: <signed-x402>" \
  -d '{
    "model": "bytedance/seedance-2.0-fast",
    "prompt": "a red apple slowly spinning on a wooden table",
    "duration_seconds": 5
  }'

# 2. Returned id is like "vd_abc123" — poll until completed:
curl -H "x-payment: <signed-x402>" \
  "https://blockrun.ai/api/v1/videos/generations/<id>?model=bytedance/seedance-2.0-fast&duration=5"

# Keep the same character across clips: pass a ta_xxx asset id
# (enroll via Virtual Portrait or RealFace — see below)
#   "real_face_asset_id": "ta_xxxxxx"

Use the BlockRun Python SDK or TypeScript SDK to handle x402 payment and polling automatically.

Start generating with Seedance

Fund your wallet with USDC on Base and generate your first clip immediately. No registration, no API keys required.

Open in Playground View All Services

Seedance API questions: text to video, image to video, characters, pricing

Does Seedance do image-to-video as well as text-to-video?: Yes. Every Seedance tier on this page accepts a seed image to animate, as well as a text prompt.
How long can a Seedance clip be, and at what resolution?: Each tier's row states its duration ceiling and resolution; the newer line runs longer. Output defaults to the resolution on this page with synchronized audio on the text-to-video path.
How is Seedance priced through BlockRun?: Per second of generated video at the tier's rate — each tier's row shows the rate and the price of a default-length clip — quoted in the 402 before generation starts.
Can I keep the same character across Seedance clips?: Yes. Enroll a character once, as a Virtual Portrait or through RealFace, then pass its asset id on each request; the tiers that support it are marked on this page.
Does Seedance output include audio?: Text-to-video output ships with synchronized audio by default, matching what ByteDance's own product returns. Image-to-video follows the model's own behaviour, stated on its row.
Seedance via BlockRun vs Runway, Luma or fal.ai — which should I use?: Runway and Luma sell their own models on a subscription; fal.ai hosts others' behind a key and a balance. Here Seedance's tiers are paid per clip with no account, beside Sora and Grok Imagine on the same endpoint.
How do I use the Seedance video API without a BytePlus account?: Call BlockRun's video endpoint with a Seedance model. It generates the clip on ByteDance's Seedance family, priced per clip and paid by the request — no BytePlus or JiMeng account and no API key.
Can an AI video pipeline call Seedance unattended?: Yes. The request pays for itself and returns a job to poll; the finished MP4 lands on a hosted URL that does not expire, so the next step in the pipeline can take it directly.