Version 2.4.0 brought significant training updates to two of the most popular models to reduce the "uncanny valley" effect:

: For the first time, users could opt-in to early-access beta models directly through the Model Manager preferences.

: Accessible under the "File" menu, it lets you download, delete, or move models to manage disk space efficiently—critical when dealing with high-resolution AI processing. 2. Refined AI Models: Natural Over "Artificial"

: Enhanced training for better frame interpolation (slow motion), fixing artifacts to ensure motion looks more natural and less "melted".

Topaz Labs has long been the "voodoo magic" of the creative world, and the update, released in August 2021, remains a landmark version for users who prefer the legacy "Video Enhance AI" branding over the newer "Topaz Video AI" subscription-based successors.

: 16GB is the minimum for stable 1080p work, but 32GB is advisable for 4K upscaling.

: You can now update AI models independently of the main application software, meaning you don't have to reinstall the whole program to get the latest "brain" for your video.

: A dedicated GPU with at least 6GB to 8GB of VRAM is recommended (NVIDIA RTX series is the gold standard here).