Release Log
New models, canvas and workflow improvements, and everything else we ship.
LTX-2.5 is now on YouArt
LTX-2.5 is now live on YouArt with audio-driven video generation: connect an audio track and the video automatically follows its length, frame rate, and audio settings. A unified frame picker and clearer “following audio” controls simplify setup.
Midjourney v8.2 is now on YouArt
Midjourney v8.2 is now the default Midjourney model on YouArt, delivering bolder, more refined results with native 2K HD rendering, dual reference controls, and the new --weird parameter.
What’s new
Midjourney v8.2 is now the default Midjourney model on YouArt, while v8.1 remains available. It produces bolder, more refined visuals with fewer low-quality results per batch.
Core capabilities
- Native 2K HD rendering; enable --hd to receive four HD images per generation.
- Text-to-image and image-prompt generation.
- Dual reference channels: Image Prompt for content, composition, and color control via --iw; Style Reference for style transfer via --sref and --sw.
New --weird parameter
Use --weird to create deliberately unconventional results. The supported range is 0–3000, with a default of 0.
Reference note
Midjourney v8.2 does not support character or omni references. For stronger subject consistency, use an image prompt with a higher --iw value and a precise text description.
Seedance 2.5 is now on YouArt
Seedance 2.5 and Seedance 2.5 Omni Reference are now available on YouArt, combining native-audio video generation with up to 30-second outputs. Generate from text, images, frames, or referenced assets, starting at 20 credits per second. Try it now →
MiniMax H3 is now on YouArt
2K resolution with native audio. Generate video from a text prompt, a first or last frame, or up to 12 reference assets. Try it now →