Skip to main content
← Back to news

LTX-2.5: synchronized video and audio locally, 22B parameters and native multishot

Research
Source:Hugging Face
Lightricks releases LTX-2.5, open weights: video generation with synchronized audio from text, image, or video, with native multishot that holds identity and style across scenes, a new diffusion video decoder replacing the VAE stage, and a Gemma 4 12B text encoder. The weights are built for running and Fine-tuning on your own infrastructure.
Read original article

Related articles

Nemotron-Parse: NVIDIA's OCR family for reading documents locally

NVIDIA re-released Nemotron Parse (v1.1, v1.2, 2.0) in the same week, on October 20: under 1 billion parameters, page structure included. Local setup with vLLM and the honest limits.

Qwen-Image 2.1: Generate and Edit Images Locally

Qwen's 7B open model that generates, edits, and creates transparent images: how to run it locally with diffusers and ComfyUI.

Local Images: ComfyUI 0.40 and comfy_attention — new for all models

ComfyUI 0.40 extends the comfy_attention attention engine to all models, turns on the local asset API, and standardizes 3D output. Plus the VRAM table that tells you what runs on your hardware.

Get the next local-AI briefing

Once a week, the top selection from iaLocale: articles, news and resources on local AI, and more.

After signing up you receive a confirmation email (double opt-in). Without confirmation we send nothing. No spam, we promise.

Related resources

Tools and resources from the iaLocale archive relevant to this content.

Kandinsky 6.0 Video

Models

Kandinsky Lab's foundation models for synchronized video and audio generation: from a prompt they generate video clips with aligned audio, released as open weights. Top paper of the day on Hugging Face (2026-10-06), with checkpoints and Spaces already available.

⚡

faster-whisper

Inference

Reimplementation of Whisper based on CTranslate2: up to 4 times faster than openai/whisper with the same accuracy and less memory. Supports 8-bit quantization on both CPU and GPU.

💎

Gemma 4 12B

Models

La taglia 12B della famiglia Gemma 4 di Google: 11,95B dense, contesto 256K e multimodalità integrata senza encoder separati (testo+immagini+audio). Licenza Apache 2.0. In Q4_K_M occupa circa 7,4GB: il punto di equilibrio tra qualità e requisiti hardware per l'uso locale su 8-16GB.

🎵

Kokoro-82M

Models

82-million-parameter speech synthesis model on a VITS architecture, Apache-2.0 for the weights, with voices in dozens of languages and surprisingly high quality for its size. Runs efficiently on CPU alone.

Browse all resources