amige. is an AI content detection tool focused on judging whether an image, video, or text is more likely to be AI-generated or human-created, and, when confident, indicating which generative model the content most resembles being made by.
In terms of use, amige. is aimed at scenarios that require discerning content authenticity, such as identifying AI-generated images, deepfake videos, suspicious text, and more. The official description provides multiple example scenarios, including fake breaking-news images, fake dating photos, video calls with out-of-sync lip movements, rental or hotel listings with abnormally low prices, fake images of limited-edition product releases, AI works impersonating famous paintings, and comments suspected of being AI-generated.
In terms of main capabilities, amige. supports three media types: images, videos, and text. A single scan compares against more than 90 generative models and lists the most likely model names. It uses a "detector panel" mechanism: routing content to multiple independent detectors best suited to that type, with each detector giving its own result, which are then fused and calibrated into a final estimate. The official description emphasizes that results show only probabilities, not proof, and that probabilities are constrained so that 0% or 100% is never displayed. When detectors disagree or the content is anomalous, amige. marks the conclusion as "uncertain" rather than guessing. In addition, it runs deepfake-specific checks for images and videos and labels the most likely type of manipulation in the results.
In terms of how to use it, users can upload images, videos, or text, with a maximum single file size of 15MB, and HEIC and MOV formats are supported. One scan can be run without an account; the free tier allows 3 scans per day; paid plans provide higher image and video quotas and unlimited text. Detection results show the independent readings and disagreements of each detector, rather than only giving a single averaged conclusion.
In terms of applicable scenarios, amige. can be used to verify images or videos circulating on social media, check images in rental or product listings, assess the authenticity of reviews, identify suspected deepfake content, and more.
In terms of limitations, the official description states that accuracy declines on heavily edited or repeatedly forwarded content; all results are probabilities rather than evidence; the novelty check only flags content anomalies, does not claim what generated it, and is not a new-generator detector. The official description does not provide specific accuracy figures, and only states that the readings in the animation are merely illustrative of a single scan and are not an accuracy claim.
