Vance Video AI

VanceAI is a photo and video processing suite that groups a large set of enhancement and restoration models behind one account. It covers the repetitive work that sits between a raw capture and a usable asset: enlarging, sharpening, cleaning up noise, repairing damage and cutting subjects out of their backgrounds. The same jobs can be run in the browser or on the desktop. The image side includes upscaling of up to 40 times, sharpening for motion blur and focus errors, denoising, retouching, restoration of faded or scratched photographs, colorization, dedicated anime upscaling, photo to sketch and photo to cartoon conversion, background removal, image cropping and passport photo preparation. The video side adds an automated enhancement pass, upscaling, standard to high dynamic range conversion, colorization, denoising, frame interpolation for smoother motion and subtitle generation. Two desktop clients are published, one that processes locally and one that uses the cloud, and developers can reach the same models through an API, a ComfyUI node or an agent skill. The company says its models are built on deep convolutional neural networks. Subscriptions are credit based. On annual billing the published tiers are Starter at $15 per month, billed $180 annually, for 260 credits a month; Popular at $30 per month, billed $360, for 650 credits; Creator at $45 per month, billed $540, for 1,300 credits; and Studio at $75 per month, billed $900. Monthly billing and a pay as you go option are also offered, and a seven day money back guarantee applies. The desktop software is sold separately at $39.90 for one month, $99.90 for one year or $129.90 for a lifetime license. VanceAI was founded in July 2020 by a team of deep learning researchers in Hong Kong, with offices in Hong Kong and Japan.