yt-whisper

yt-whisper is a tightly scoped open-source command-line tool that uses yt-dlp to download a YouTube video and Whisper to transcribe it, producing a SubRip file. The pitch is simplicity. You install the package, run one command with a YouTube URL, and an SRT file lands in your working directory.

It is the easiest possible bridge from a public video to a usable caption track, and it pairs naturally with larger pipelines that also use yt-dlp for archival or processing. There is no GUI, no styling, no burn-in.

Output is plain SubRip suitable for upload back to YouTube, ingestion into a styling tool, or use in any application that consumes standard subtitle files. The dependency on Whisper means local compute determines speed — with a GPU, transcription completes in a fraction of real time, and on CPU only it runs slower.

Multi-language support comes from Whisper directly, which means transcription works across most major languages with auto-detection or explicit language flags. For developers building captioning workflows on top of YouTube content, yt-whisper is a useful building block. For end users without command-line comfort, it is the wrong shape — pair them with a wrapping tool or send them to a hosted alternative.