Details
The platform's core image creation tools (Splicer, Splicer 2, Collager, Composer) are built on generative adversarial networks (GANs), specifically StyleGAN and BigGAN, which take one or more images and user-adjusted trait parameters as input and produce new synthetic images as output. Users can crossbreed two or more images from the shared public library, upload their own images for the model to encode, and fine-tune more than 30 visual attributes such as age, gender, skin tone, color balance, and facial features. A freemium model provides basic access to all tools for free users (with limited monthly credits), while paid subscribers unlock higher resolution exports, private mode, and additional generation capacity.
Have evidence about Artbreeder's AI practices? Submit a report.
Report a Sighting →