Kandinsky 6.0 video gen + video upscaler!
Summary
Kandinsky Lab releases Kandinsky 6.0 Pro (29B) and Lite (3B) video generation models with a built-in video upscaler, already supported in ComfyUI and Diffusers.
Similar Articles
Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation
Kandinsky Lab presents Kandinsky 6.0 Video, a family of foundation diffusion models (Lite 3B and Pro 29B) for synchronized text-to-audio-video and image-to-audio-video generation with lip-synced 44kHz audio and Full-HD super-resolution. The models use a dual-stream CrossDiT architecture with bidirectional cross-attention, outperform Kandinsky 5.0, and are released with code, checkpoints, and diffusers integration under the MIT license.
LBH-123-AI/Minimax_h3_latent_Upscaler
A neural latent-space upscaler designed to accelerate high-resolution Minimax H3 video generation by upscaling latent representations directly, avoiding expensive decode-encode round-trips.
Comfy-Org/MiniMax-H3
Comfy-Org repackaged MiniMax-H3 model files for ComfyUI, including diffusion models, text encoders, and VAEs, with workflow templates for text-to-video, image-to-video, and reference-to-video generation.
MiniMax H3 Day-0 Support in ComfyUI: Open Weights, Native Audio, and 2K Video
MiniMax H3, a next-generation open-weights video model capable of generating 2K video with native stereo audio from text, images, video, or audio, launched with day-0 ComfyUI support and optimizations that allow it to run on consumer GPUs.
Lightricks/LTX-2.5
Lightricks releases LTX-2.5, an open-weights world model for generating synchronized video and audio from text, image, and video inputs, with features like native multishot generation and a new diffusion video decoder.