AI PDF Reader – Active Reading Companion is a tool designed to assist users in reading and understanding dense PDF documents. Its primary function is to act as a page-by-page reading companion that interprets the content of each page using a vision model. The tool renders each PDF page as an image, allowing the model to read text, equations, tables, figures, scans, and multi-column layouts exactly as they appear on the page. Users can interact with the companion via voice or text, and it provides explanations, clarifications, and encouragement as they progress through the document. The main features include the ability to handle a wide variety of document types, such as academic papers, textbooks, specifications, contracts, and stories. The vision model processes the visual representation of each page, so it sees the same content the user sees, including complex formatting. Interaction can occur through voice or text input and output, making the experience adaptable to user preference. The tool operates at the user's own pace, offering support page by page rather than requiring the user to read the entire document alone. Typical use cases involve reading materials that are difficult to get through without external help. This includes research papers, dense textbooks, legal contracts, or any document where understanding every detail is important. The workflow begins with uploading a PDF, after which the companion walks through each page, answering questions and providing explanations as needed. The user can choose to read privately with the companion, similar to having a friend or advisor available to explain difficult sections. Technically, the service runs and funds its own models, so users do not need to manage API keys or worry about per-token costs. The first 50 pages are free to read without requiring a credit card or setup. After that, a simple subscription unlocks unlimited reading. The tool was originally developed to help with machine learning and math papers read alone late at night, and it has since been generalized to any document that requires deep understanding.