v0.1.0-alpha

0.1.0-alpha - Codename: Frame

This is the very first version, experimental MVP, of Vikit.ai SDK.

It was originally private source and allows generating videos from an audio recording (e.g. a blog audio) or text. It also allows for merging the final video with background music and read aloud the prompt generated from text.

Music can be generated, or a default royaltee free background music. You may also use your own prompt recording as an audio track.

0.1 uses various models for video generation, music generation, audio prompt to text and synthetic voice generation.

Available type ov videos are:

• RawTextBasedVideo: the most basic building block that usually corresponds to a 2 to 6s video , depending on the generative model and platform providing it
• CompositeVideo: A video made of other video. You will probably play a lot with this beast and the RawTextbasedVideo!
• PromptBasedVideo: for lazy people like us 😉 who just want to give a prompt and let the video be generated. it inherits composite video and includes a first simple way to stitch video and transitions together
• Imported (i.e. your video, generated or not)

Initial beta testers experimented with 0.1 using a Google Colab, though video could take a long time to generate if you stacked many scenes within (tens of minutes).