know-fast

fast is a tool that processes links to articles, videos, podcasts, and papers, returning structured knowledge outputs for each submitted URL. For a flat fee of one cent per processed link, it provides a summary, a set of graded and cited claims, and the full speaker-labeled, timestamped transcript where applicable. The service is designed to deliver a concise, approximately 2,000-token knowledge packet that can be instantly loaded by compatible agents, enabling rapid access to distilled information.

The platform supports a range of content types, including YouTube, Spotify, Podcasts, X, Substack, and academic papers. fast generates speaker-labeled, timestamped transcripts, while articles are returned as clean full text. Every processed link includes graded evidence in a structured format. Once a link is processed, the knowledge packet becomes instantly accessible for all users at the same fixed price, eliminating repeated processing costs for the same content.

Integration with agents is facilitated through a setup prompt that can be used with platforms such as Hermes and OpenClaw, allowing agents to automatically convert encountered links into knowledge packets. The tool is positioned for users who need to extract, organize, and reference information efficiently from diverse media sources.

01 per processed link. There is no mention of a subscription model or open-source licensing. The service handles the entire processing pipeline, including caption fetching, speaker labeling, claim extraction and grading, and report synthesis, and delivers the results in formats such as summary, key findings, graded claims, and transcript files.