Upload an image (a screenshot, a quote, a product photo). A vision model selected by our router describes it and extracts text, grounded and without identifying specific people. We mask the question text; the image is sent to a cloud model, processed temporarily, and not retained by the application.
Sample prompts:
Image up to 5 MB. No person identification. We mask the question text; the image is sent to a cloud model, processed temporarily, and not retained by the application.