VisionMate
Describes images with AI, locally on your computer. Describe the current element, the whole screen or a clipboard image. Press once for a quick spoken description, twice for a detailed one in a window you can copy from. Automatic descriptions of unlabeled images are optional and off by default.
After a one-time model download of about 1.5 GB it works offline, with no account, no API key and no cloud upload. Requires Windows 10 or 11 (64-bit) and NVDA 2024.1 or later. The source is publicly available under GPL v2 or later.
What it does for you: spoken descriptions of images, with images never leaving your computer.