OpenScribe Studio - Free Video, Audio, and YouTube Transcript Generator

OpenScribe Studio is a free, open-source desktop app for turning videos, audio files, and YouTube/video links into readable transcripts and subtitle files.

It is built for normal people first: creators, students, researchers, teachers, editors, marketers, writers, and anyone who needs a transcript without paying per minute to a cloud service.

OpenScribe Studio runs speech-to-text AI on your computer using Whisper through faster-whisper. Local files stay on your device. YouTube and other video links still need internet because the app must download the audio before it can transcribe it.

Project History

OpenScribe Studio started as a private tool in December 2025 for my own transcription workflow. I used it to turn videos and audio into transcripts that could become notes, subtitles, research material, and source context for AI agents.

After improving the desktop interface, progress tracking, cancel support, documentation, and release builds, I published it as an open-source project on May 26, 2026.

What You Can Do With It

• Create transcripts from MP4, MP3, WAV, MOV, MKV, and many other media files.
• Paste YouTube or video links and turn them into transcript files.
• Export subtitles for video editing or upload: SRT and VTT.
• Save readable notes as Markdown or plain TXT.
• Save structured JSON for automation, search, datasets, or AI workflows.
• Translate speech into English using Whisper translation mode.
• Run locally with no per-minute transcription fee.
• Use CPU by default, or NVIDIA CUDA GPU if your computer supports it.

Why Use OpenScribe Studio?

Many transcription tools are paid, online-only, or locked behind monthly plans. OpenScribe Studio is different:

• Free and open source: use it, study it, modify it, and share it.
• Local-first: your local files are processed on your computer.
• Beginner-friendly: a desktop interface, progress bar, ETA, and cancel button.
• Useful outputs: transcripts, subtitles, plain text, and JSON.
• Good for AI work: turn videos into clean text that can be used as research notes, knowledge bases, or training material for agents.

Example Use Cases

OpenScribe Studio is useful when you want to:

• transcribe YouTube tutorials into study notes
• create subtitles for your own videos
• turn lectures, meetings, podcasts, webinars, or interviews into searchable text
• extract quotes and timestamps from long videos
• archive family recordings or voice notes
• create written summaries from course videos
• prepare source material for blogs, newsletters, and scripts
• build datasets from spoken content
• create training material for AI agents from video transcripts

One real workflow: download educational or strategy videos, extract transcripts, clean the text, then use those transcripts as source material for AI agents that need to learn a specific process, style, domain, or workflow.

Download

Go to the latest release:

https://github.com/michaelxer/openscribe-studio/releases/latest

Download the file for your system:

• Windows: OpenScribeStudio-windows-x64.zip
• macOS: OpenScribeStudio-macos.zip

The Windows version is portable. Unzip it, open the folder, and run OpenScribeStudio.exe.

The macOS version is currently unsigned. It is useful for testing, but macOS may show a security warning until the app is signed and notarized with an Apple Developer account.

Quick Start

1. Open OpenScribe Studio.
2. Add local video/audio files, add a folder, or paste one YouTube/video URL per line.
3. Choose an output folder.
4. Leave Language as Auto detect unless you already know the spoken language.
5. Start with these safe settings:

Model: small
Device: cpu
Compute Type: int8

6. Select output formats such as MD and SRT.
7. Click Start transcription.

The app shows progress, elapsed time, estimated time remaining, current activity, and a Cancel button.

Which Model Should I Choose?

If you are unsure, use small.

Model	Best for	Notes
tiny	Fast test runs	Lowest accuracy
base	Quick drafts	Better than tiny, still basic
small	Most people	Best default balance
medium	Better quality	Slower, uses more memory
large-v3	Best quality in this app	Slowest, best for difficult audio
distil-large-v3	Fast English transcription	Best for English, not ideal for translation

Simple recommendation:

• Normal laptop or desktop: small, cpu, int8
• Weak computer: base, cpu, int8
• Important transcript: medium or large-v3
• NVIDIA GPU with CUDA working: medium or large-v3, cuda, float16
• English-only and speed matters: distil-large-v3

For more detail, read USER_GUIDE.md.

Output Formats

Format	Use it for
MD	Easy-to-read transcript with timestamps and metadata
SRT	Subtitles for YouTube, editors, and media players
VTT	Web subtitles
TXT	Simple plain text
JSON	Automation, datasets, search, and AI pipelines

Supported Files

Local files can be audio or video. Supported extensions include:

3gp, aac, aiff, avi, asf, flac, flv, m4a, m4v, mkv, mov, mp3, mp4,
mpeg, mpg, m2ts, m2v, mts, mxf, oga, ogg, ogv, opus, ts, wav, webm, wma

Privacy

OpenScribe Studio is local-first:

• Local files are transcribed on your computer.
• The selected Whisper model may be downloaded the first time you use it.
• YouTube/video URLs require internet because the audio must be downloaded.
• The app does not include a chatbot LLM.
• The app does not charge per minute.

Install From Source

This option is for developers or advanced users.

Requirements:

• Python 3.10 or newer
• ffmpeg
• Windows, macOS, or Linux with Tk support

Install dependencies:

python -m pip install -r requirements.txt

Run the desktop app:

python transcript_app.py

On Windows:

.\run_app.ps1

Build Windows Portable App

.\build_windows_portable.ps1

The portable build is created at:

dist\OpenScribeStudio\OpenScribeStudio.exe

Build macOS App

On macOS:

chmod +x build_macos_app.sh
./build_macos_app.sh

The app bundle is created at:

dist/OpenScribeStudio.app

For public distribution outside the App Store, macOS users get the best experience when the app is signed and notarized.

Command-Line Examples

Transcribe one local video:

python transcript_tool.py --input "D:\Videos\meeting.mp4"

Transcribe every supported media file in a folder:

python transcript_tool.py --input "D:\Videos" --model medium --formats md srt json

Transcribe YouTube links from a text file:

python transcript_tool.py --urls-file urls.txt --output-dir transcripts

Transcribe Indonesian audio and translate it to English:

python transcript_tool.py --input video.mp4 --language id --translate --formats md srt

Keywords

Free transcription software, free video transcript generator, YouTube transcript generator, audio to text app, video to text converter, free subtitle generator, SRT generator, VTT subtitle tool, Whisper desktop app, local speech-to-text, open-source transcription tool, podcast transcription, lecture transcription, meeting transcription, AI training transcript tool.

Roadmap

• Automatic PC capability recommendation for model/device settings.
• Signed Windows installer.
• Signed and notarized macOS DMG.
• Better per-download progress reporting for every supported video site.
• Screenshots and short demo video.

License

MIT. See LICENSE.