ModelScope Text-To-Video
A text-to-video model with downloadable weights for local use; it accepts English input only. Its license restricts commercial use.
Description
ModelScope Text-To-Video is an open-weight model that generates short video from a text description. Its Hugging Face card provides files and Diffusers setup instructions.
The hosted ModelScope demo is marked stopped. The model can still be downloaded and run locally with suitable hardware.
The model card specifies English input only. License terms restrict commercial use; check the full model card before using it.
Sources and review (2026-11-02): https://huggingface.co/ali-vilab/text-to-video-ms-1.7b; https://modelscope.cn/models/AI-ModelScope/modelscope-damo-text-to-video-synthesis; https://modelscope.cn/studios/damo/text-to-video-synthesis.
