Leonardo.Ai launches the Pro Upscaler

Leonardo.Ai launches Pro Upscaler, a proprietary tool scaling AI-generated images up to 105 megapixels via Precise and Creative modes integrated with Canva.

Leonardo rolls out the Pro Upscaler, an in-house developed image upscaling tool with an architecture designed exclusively for AI-generated content. Unlike traditional upscalers, trained on real photographs, this tool incorporates a proprietary pipeline capable of recognizing synthetic textures, artificial grain, and characteristic artifacts of AI models, which traditional systems tend to amplify rather than correct. The maximum resolution reaches 105 real megapixels: a 5,100 x 5,100 pixel image can thus be scaled up to 10,200 x 10,200 pixels in a single pass. This threshold covers nearly all professional uses, from large-format display to billboard printing, including 8K displays and full-page magazine spreads. The tool comes in two distinct modes. The Precise mode enlarges the image without altering its composition or introducing generated elements, with an optional AI artifact correction parameter. The Creative mode activates a generative reconstruction of the image during the upscaling process, useful when the source presents structural defects such as malformed hands or distorted faces in the background. This second mode offers three intensity levels: Subtle, Balanced, and Strong. A multiplier, adjustable from 2x to 8x, determines the extent of the applied processing. The Pro Upscaler is directly accessible from the Leonardo interface, via the dedicated tab or from the generation feed. An integration with Canva allows for exporting final files without leaving the platform. The tool is available on all active accounts and billed in tokens, depending on the selected mode and multiplier.