Vvideo4

Seedance 2.0 connected

AI content generation
Web studio / Open API both ways

Generate right in the browser, or wire the same capability into your own product with a single key. One credit pool, one bill, pay for what you use.

Web studio

A studio that works out of the box. The asset library keeps reference images and finished takes in one place; characters keep the same person consistent across many generations.

  • · Cost is estimated before you generate — no nasty surprises
  • · Multiple reference images, reference video and reference audio
  • · Results are stored on our side, so expired upstream links never lose your work

API service

Embed generation into your own site or tooling, or resell it as SaaS. Every key gets its own rate limit and monthly cap, so you can carve up quota per customer.

  • · BytePlus ModelArk compatible — swap the base URL and your code just works
  • · A unified REST layer that hides vendor differences
  • · Failed tasks are refunded in full, automatically

Available models

More are being added; new models need no changes on your side

Seedance 2.0API ready
seedance-2-0

Flagship. 480P/720P/1080P/4K, multiple reference images, reference video and audio

Seedance 2.0 FastAPI ready
seedance-2-0-fast

Fast tier. Quicker output at about 80% of flagship cost. 480P/720P

Seedance 2.0 MiniAPI ready
seedance-2-0-mini

Lightweight tier. Cheapest, about half of flagship. 480P/720P

Seedream 5.0 ProAPI ready
seedream-5-0-pro

Top image model for characters and scenes. Output works directly as video first/last frames or reference images

Seedream 4.5API ready
seedream-4-5

Previous-generation image model. Cheaper, and good enough for everyday use

Seed-SCAPI ready
seed-sc

Role-play model that writes character-driven scripts and video prompts. Requires upstream allowlisting

Billing

Billed per token, on the same basis as the upstream invoice: token count follows resolution and duration, priced per million tokens. Settlement runs on credits, 1 yuan = 1 credits. The web studio and API calls share one balance, and the bill breaks down each. Failed tasks are refunded in full, automatically.