Art generated from words, or grown from an image.
NEUART.ai was our creative AI label in 2022, a showcase rather than a product for sale. It worked in two ways: pure AI, where images, video and animation come straight from text, steered with words, phrases and concepts; and image-based, where an existing picture or video is the seed of a one-off piece that cannot be made again.
Behind it we built our own generation service: the open-source engines, Disco Diffusion, Stable Diffusion and pytti, taken out of notebooks and packaged in a container with a small API, one call for text to image and one for image or video to video, with the camera moving through the scene frame by frame. It grew out of the work for Nefty, and it is what NEUART used to make its pieces.

Describe
Words, phrases and concepts steer the model.

Generate
Images, video and animation come straight from the text.

Seed
Or an existing picture or video becomes the starting point.

Serve
A containerised API runs the engines and returns the result as a link.



Working on something like this? Tell us what you need.

