LTX-2.5 Video Model Released: Features Multishot Scenes and High Fidelity

linoy_tsaban · x · 2026-08-12

Lightricks introduces LTX-2.5, a new video generation model emphasizing multishot scene consistency and high pixel fidelity. It incorporates Diffusion Fidelity Rendering to ensure cinema-quality frame-by-frame output.

Designed as an open foundation, the model along with quantized versions and accompanying LoRAs has been released on Hugging Face, supporting fine-tuning across various domains.

Related event: LTX-2.5 Open-Source Video Model Released: Native Multishot and Fast Inference(20 posts)→

Original post →

More from Multimodal

Multimodal channel →