still needs to wait for open source to develop text to video models with audio generation. If they have image to video with audio they won't take that long to release the new models since they already has all the previous text to video models already, and video to video etc