🐰 Mochi — Free YouTube Downloader + AI Subtitles & Transcript

Tools & Scripts
, ,

AI-generated summary

:bullseye: What you get: Turn a YouTube link into high-quality video, audio, timed subtitles, and plain text transcripts on your machine.

:white_check_mark: Start here:

  • Mochi — downloads media and runs offline AI speech recognition.
  • First-time installer — grabs dependencies and downloads the local AI transcription model.

:toolbox: Also useful:

  • Subtitle exporter — creates synced .srt and .vtt caption files.
  • Audio rip — extracts original sound or MP3 tracks.

:warning: Watch out:

  • Initial setup needs an active internet connection to download the required tools and AI model.
1

One link in — four files out

:link: one YouTube link → :clapper_board: video · :headphone: audio · :speech_balloon: subtitles · :memo: text

Mochi — paste a link, pick an output
Mochi — paste a link, pick an output1536×1075 58.3 KB

One YouTube link → pick the output → Make it happen ✦

:rabbit_face: Mochi turns a single YouTube link into four useful files — free, on your own computer, for Windows, macOS and Linux.

You pick You get
:clapper_board: Full video Video + audio, best available quality
:headphone: Audio only Original audio or MP3
:speech_balloon: Subtitles .srt + .vtt, timed to the speech
:memo: Transcript .txt of everything said

The rare bit — subtitles and transcript come from AI listening to the real audio on your machine, so you get captions even when the video has none. Works without an account and without any cloud speech service.

latest release · Windows · macOS · Linux

Under the hood — :rabbit_face: Mochi just wires four well-known tools together:

How Mochi works — one YouTube link, two paths, everything saved to your Downloads
How Mochi works — one YouTube link, two paths, everything saved to your Downloads2748×2080 260 KB

First run — one file per system

1. Extract the ZIP and open the Mochi folder.
2. Run the installer for your system:

System First time / update Every day after
:window: Windows Install-Mochi-Windows.cmd Run Mochi Now.lnk
:red_apple: macOS Install-Mochi-Mac.command Run Mochi Now.app
:penguin: Linux Install-Mochi-Linux.sh Run Mochi Now.desktop

3. Paste the link → pick the output → Make it happen ✦ → Downloads / Open my saved files.

First run sets everything up · after that, one click opens it

First run uses the internet to install the tools and fetch the AI model; after that, speech recognition runs on your own machine. Grab videos you have permission to save. Built-in Fix my tools sorts out known tool issues, and your saved files are kept across updates.

More → source + full guide on GitHub · report an OS bug or idea at Issues

Made with :purple_heart: by 1Hack — tell us which OS you run it on

1 Like
2

Local transcription is the part I like most here, since videos without captions are usually the ones worth saving. I would test it first on a noisy video with a strong accent, because that is where automatic captions usually break down. Checking whether the timestamps drift on a longer video would also tell me if the subtitles stay usable for study notes. Keeping the plain text transcript next to the audio makes it easy to search later without replaying the whole video. For full downloads, I would start with one short video to see how much disk space the best quality setting really uses.