Home / Technology / ByteDance Unveils Seedance 2.5 AI Video Model
ByteDance Unveils Seedance 2.5 AI Video Model
24 Jun
Summary
- Seedance 2.5 generates up to 30-second videos, doubling predecessor's limit.
- Users can now include 50 reference materials in a single prompt.
- New model ships next month in China, global launch uncertain.
ByteDance introduced its latest AI video generation model, Seedance 2.5, during a conference in Beijing. This new model boasts substantial upgrades over its predecessor, Seedance 2.0. The company announced plans to release Seedance 2.5 in China next month.
Seedance 2.5 offers users enhanced capabilities, including the generation of videos up to 30 seconds long, a notable increase from the previous 15-second limit. Furthermore, the model supports the inclusion of up to 50 reference materials—such as images, videos, and audio files—within a single prompt, greatly improving creative precision.
This launch follows the controversial release of Seedance 2.0 earlier in the year. Users generated viral videos using copyrighted characters, leading to legal disputes with major Hollywood studios. To mitigate future issues, BytePlus implemented restrictions on realistic human faces in prompts.
The global availability of Seedance 2.5 is currently uncertain, contrasting with its confirmed rollout in China.
