MongoDB - Build AI That Scales
Mini toolsLTX 2.3
By Lightricks
LTX-2.3 is an open-weights diffusion model for joint audio-video generation (one model produces both modalities), built around the core capabilities of modern video generation and shipped with multiple checkpoints, including a full 22B “dev” model, an 8-step distilled model, a LoRA variant, plus latent spatial and temporal upscalers for higher resolution and higher FPS pipelines.
New
Multimodal
Gen 3
Overview
LTX-2.3 is a DiT-based audio-video foundation model that generates synchronized video and audio in a single model, with improved audio-visual quality and stronger prompt adherence versus LTX-2.
About Lightricks
Use all-in-one photo editor Photoleap for amazing creations on your iPhone: Change backgrounds, remove objects, create collages, apply filters & effects. Try a 7 day free trial now.
Location:
Jerusalem, IL
Website:
photoleapapp.com
Tools using LTX 2.3
-
LTX by Lightricks — v2.3Joint audio-video generation in a single model, producing synchronized video and sound together. Supports multiple generation modes including text-to-video, image-plus-text-to-video, audio-to-video, and text-to-audio-video. Released with open weights and a focus on practical local execution. Includes both a full trainable model and a distilled version for faster inference, plus spatial and temporal upscalers. Supports local and developer workflows through Diffusers, ComfyUI, and the official PyTorch codebase.
Last updated: March 5, 2026
