Changelog

Follow up on the latest improvements and updates.

RSS

  • We have completely refreshed the UI
  • The base model is now SDXL
  • It now generates images with the resolution of 1024x1024 px instead of 512x512
  • You can now download generated images
  • There is a new set of styles now. We removed the old ones and added new ones instead: Base, Anime Illustration, 3D Toon, B&W Sketch
  • We removed the Image2Image function
Image to Video allows you to add motion to any image. You can upload an image and transform it into a short video.
  • supports 16:9, 9:16 and 1:1 aspect ratios
  • from 4 to 24 FPS
  • Added Essential pipeline to API with all 3 styles: Photorealism, Artistic, and Anime.
  • lowered the prices of API by about 4x on average.
We’ve added a new simplified UI to the AI Generator, called
Essential
. It greatly changes what was previously possible with the AI Generator. Thanks to our new custom pipelines and fine-tuned models under the hood, it has increased the default quality of images generated with simple prompts while increasing the generation speed.
The previous functions are in the
Advanced
tab. In the
Essential
tab, there are the key features needed to generate high-quality images.
  • You can now upload your pictures and change their style or edit them using prompts.
  • There’s also a strength parameter, just like in the AI Generator.
  1. We've added an option to select styles. You can now choose between Photorealism, Art, and Anime. All 3 styles are, in fact, our custom in-house trained models.
  2. Real-Time got even faster
Real-Time AI Generator allows you to generate images in real-time, so while writing a prompt, the image changes before your eyes.
  • No more smoothing your creations
    - The old Upscaler had a tendency to smooth out images. This issue is fixed with the new one. You can now make the image 4x bigger and refine every single pixel with outstanding effects.
  • 4x Upscale
    - You still get to upscale your images 4x, up to 6144x6144 px. But the quality is now much better.
  1. We have added a Strength parameter to ControlNet
  2. We have added 3 new ControlNets:
  • Brightness - can be used to turn B&W images into color or any image recoloring/restyling
  • QR Codes v2 - can generate scannable QR Code resembling the prompt
  • QR Code Monster v2 - same as above but can also be used with any pattern allowing infinite creativity
We’ve added 10 new Stable Diffusion XL-based models to the AI Generator. These are:
  • RealVisXL V1.0 - produces realistic images
  • InfiniAnimeXL - anime style images
  • DynaVision XL - produces stylized 3D model output similar to computer graphics animation like Pixar, Dreamworks, Disney Studios, Nickelodeon, etc.
  • ProtoVision XL is a portrait model, it favors showing human subjects.
  • SDVN7-NijiStyleXL - it can create a lot of great styles when combining anime, sketch and realistic styles together. You can use a combination of style art supported by SDXL.
  • XXMix_9realisticSDXL - produces realistic images
  • Lah Mysterious - aims to create fantasy atmosphere that leans more towards Asian culture, either from ancient times or an urban future setting.
  • Reproduction v3.31 - anime style images
  • Copax TimeLessXL v5 - produces realistic images
  • CounterfeitXL - anime style images
Load More