Nai offers a range of custom-trained models made by the imgnAI Team using Stability AI's open-source Stable Diffusion
Our base model, /nai, is trained on the most popular anime art styles. This model is optimized for fun and flexibility so you can input detailed or basic prompts and generate a near-infinite range of beautiful, anime-inspired art.
Our /real model lets you create photo-quality realism with incredible prompt responsiveness. From photo-realistic portraits to believable animal hybrids, make your imagination a reality with our /real model.
/hyper was built to bridge the gap between digital renders and photo-realism. From detailed 3D renders to lifelike realism, hyper has all the rizz.
Our cinematic model is trained on rich textures and cinematic lighting. This model is among the first in the world to produce perfect dark, and is capable of stark lighting contrasts. We recommend selecting “wide” image dimensions with /cin for best results.
Our /art model is designed to produce detailed images with a high degree of artistic flair. Include your favorite artist or art style in the prompt for best results.
A community favorite, /ani is trained on classic anime and produces lush, detailed anime images. Create anime art or bring the waifu of your dreams to life with our /ani model.
Derived from our realism training set, /cgi produces highly detailed digital renders with a high degree of prompt responsiveness.