---
title: "Doubao video generation model Seedance 2.0 launched"
type: "News"
locale: "zh-CN"
url: "https://longbridge.com/zh-CN/news/275699061.md"
description: "Doubao announced today that the Doubao video generation model Seedance 2.0 has officially been integrated into the Doubao App, desktop version, and web version. ByteDance's official WeChat account stated that Seedance 2.0 is still far from perfect, and its generated results still have many flaws. In the future, they will continue to explore the deep alignment between large models and human feedback to create more efficient, stable, and imaginative audio and video production tools to serve more creators"
datetime: "2026-02-12T05:29:09.000Z"
locales:
  - [zh-CN](https://longbridge.com/zh-CN/news/275699061.md)
  - [en](https://longbridge.com/en/news/275699061.md)
  - [zh-HK](https://longbridge.com/zh-HK/news/275699061.md)
---

> 支持的语言: [English](https://longbridge.com/en/news/275699061.md) | [繁體中文](https://longbridge.com/zh-HK/news/275699061.md)


# Doubao video generation model Seedance 2.0 launched

On February 12th, Doubao announced that the Doubao video generation model Seedance 2.0 is now officially integrated into the Doubao App, desktop version, and web version.

By opening the Doubao App dialog, selecting the newly added "Seedance 2.0" entry, and entering relevant prompts, users can generate 5-second or 10-second videos. Users can also choose "Avatar Video," which allows for real-person verification to create their own video avatar and experience more creative gameplay.

The Seedance 2.0 model supports original sound and image synchronization, multi-camera long narratives, and multi-modal controllable generation. By inputting a segment of prompts and reference images, users can generate multi-camera video content with a complete native soundtrack. The model can automatically analyze narrative logic, and the generated shot sequences maintain a high degree of consistency in characters, lighting, style, and atmosphere.

Currently, Doubao Seedance 2.0 does not support uploading real-person images as subject references.

ByteDance's official WeChat account stated that Seedance 2.0 is still far from perfect, and its generated results still have many flaws. In the future, they will continue to explore deep alignment between large models and human feedback, creating more efficient, stable, and imaginative audio and video production tools to serve more creators

## 相关资讯与研究

- [UTA Agent Paul Coggiola Exits to Head New Creators Division at LIFT Sports Management](https://longbridge.com/zh-CN/news/281209628.md)
- [Eightco holds $326M in treasury, heavily weighted toward AI via its exposure to Worldcoin and OpenAI](https://longbridge.com/zh-CN/news/281570584.md)
- [Deutsche Bank  Keeps Their Buy Rating on Aumovio SE (AMV0)](https://longbridge.com/zh-CN/news/281386916.md)
- [08:00 ETAIRSEEKERS Tron and Tron SE Series Now Available: Redefining "Real Mowing" with FlowCut™ Technology](https://longbridge.com/zh-CN/news/281634627.md)
- [Eurofins Scientific SE: Disclosure of Total Number of Voting Rights and Number of Shares in the Capital at 31 March 2026](https://longbridge.com/zh-CN/news/281626739.md)