跳转到主要内容

Volcano Engine Releases Seedance 2.5 API, Comprehensive Upgrade of Video Generation Capabilities

Volcano Engine officially released the Seedance 2.5 API service today. Compared to the previous generation Seedance 2.0, the new version has been upgraded in multiple dimensions such as instruction following, long-form narratives, realism, and audiovisual quality, marking another significant step forward in AI video generation capabilities.

Seedance 2.5 natively supports direct output of 30-second videos and can reference up to 50 multimodal assets, meaning creators can more flexibly combine images, text, audio, and other elements to generate richer content. Video editing features have also become more precise and stable, which is undoubtedly good news for users who need to repeatedly adjust details.

In addition to functional enhancements, Seedance 2.5 has further optimized picture quality, sound quality, lighting, camera movement, and overall aesthetics. In Volcano Engine's words, AI-generated content is increasingly approaching cinematic long-form narrative effects. Imagine that in the future, you only need to input a description to get a short film with exquisite visuals and realistic sound—isn't that cool?

Image

Volcano Engine stated that from enterprise practices, the realism and visual quality of video generation are important foundations for expanding commercial applications. The upgrade of Seedance 2.5 will bring more possibilities to scenarios such as creative advertising and industrial simulation. For example, advertising agencies can use it to quickly generate multiple creative versions, and factories can use it to simulate production processes. Things that previously required significant manpower and resources may now take only a few minutes.

Currently, Seedance 2.5 has been simultaneously launched on platforms such as LibTV, Qixiang AI, TapNow, Punch AI, DreamWorks AI, and Xingyu Film Field, making it more convenient for enterprises and creators to use these tools for AI video creation.

Additionally, Seedance 2.0 mini and Seedance 2.0 fast are also offering limited-time promotional activities. From August 7 to September 7, Seedance 2.0 mini is billed at 40% of the standard price, with a minimum of approximately 0.2 yuan per second at 720P resolution; Seedance 2.0 fast is billed at 75% of the standard price, with a minimum of approximately 0.6 yuan per second at 720P resolution. This pricing strategy further lowers the barrier to AI video creation, allowing more people to try generative video technology.

Overall, the release of Seedance 2.5 not only improves the quality and functionality of video generation but also makes it more affordable through price discounts. For content creators, advertising professionals, and even ordinary users, this is noteworthy news.

Key Points

  • Seedance 2.5 API officially launched: Comprehensive upgrades in instruction following, long-form narratives, realism, and audiovisual quality.
  • Native support for 30-second videos: Can reference up to 50 multimodal assets, with more precise and stable video editing.
  • Simultaneous launch on multiple platforms: Available on LibTV, Qixiang AI, TapNow, and other platforms.
  • Limited-time discounts: Seedance 2.0 mini and fast versions offer 40% and 75% discounts respectively, lowering the barrier to entry.