Topaz Labs Releases Video AI 7.0 Featuring Starlight Mini for Local Processing

Topaz Labs has announced the release of Video AI 7.0, introducing Starlight Mini, a new diffusion-based AI model designed for local video enhancement on Windows systems with NVIDIA GPUs.

This is the first time a diffusion model for video upscaling and restoration can run entirely on a user’s machine, eliminating the need for cloud-based processing.

Starlight Mini is optimized for handling difficult video sources such as 8mm and 16mm film, mini-DV, and interlaced footage, delivering high-quality results at resolutions up to 4K.

It currently functions as a standalone enhancement model, with future updates expected to expand its integration with additional filters.

System requirements for local use include:

Windows operating system

NVIDIA GPU with at least 8GB of VRAM (16GB recommended)

Minimum 12GB of system RAM

Local processing is not yet supported on macOS, AMD GPUs, or Intel GPUs, though these platforms can still use cloud rendering.
 

Video AI 7.0 also includes several other improvements, such as live export previews, improved EXR format support, and updated administrative controls for Pro accounts, aimed at improving workflow efficiency.

Download Topaz Video AI 7.0

For full details visit the official Topaz Labs community page.