Txtify is a web-based transcription tool for audio and video. It converts uploaded files or a YouTube URL into text, using AI and the stable-ts library for transcription. The page describes it as a free service and says it is meant to handle audio and video in multiple languages, with translation available for transcriptions.
Its input options include a YouTube URL or an uploaded audio or video file from a device. The stated file limits are up to 100 MB for uploads and up to 10 minutes for YouTube videos. The interface also lets a user choose a speech-to-text model from Whisper Tiny, Whisper Base, Whisper Small, Whisper Medium, and Whisper Large, each labeled with different tradeoffs in speed, accuracy, and language support. Language selection is part of the workflow, and translation can be enabled for the transcribed text.
Txtify also offers export in multiple formats, including .txt, .pdf, .srt, .vtt, and YouTube Captions (.sbv). The translation option includes a path through DeepL, which requires a DeepL API key and is marked as recommended in the interface. The site says this version is a demo intended to showcase the functionality of Txtify, and that the full features, including real data transcription, require self-hosting. It also points to a GitHub repository for instructions on running the application locally.
The entry identifies Txtify as an AI transcription application rather than a general-purpose editor or player, and its content centers on converting audio and video to text, translating transcriptions, and exporting the results in text and subtitle formats.
In the AI space, Txtify takes a focused approach. It focuses on transcribing audio and video files to text quickly and accurately using AI. Txtify is a consumer product aimed at users needing fast AI-powered transcription of audio and video. Txtify is free to use. Txtify is available on the web, and it can be self-hosted.
Txtify first shipped in 2024. The project is developed in the open on GitHub with 136 stars. Among its 6 catalogued features are audio transcription, video transcription, and youTube support.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do