Repository navigation
Releases
Notra v0.1.0-beta
Beta / testing release. First standalone Linux build. Expect rough edges, and please report anything that breaks.
Record a lecture, get a typeset PDF of notes. Transcription runs locally on your own machine; only the text transcript is sent to the AI provider.
Install
Download Notra-x86_64.AppImage below, then:
chmod +x Notra-x86_64.AppImage
./Notra-x86_64.AppImageThat is the whole install. Python, ffmpeg and the LaTeX engine are all inside the file.
If it will not start, install the library AppImages use to mount themselves:
sudo apt install libfuse2Or skip that entirely:
./Notra-x86_64.AppImage --appimage-extract-and-runRequirements
| OS | Ubuntu 22.04 or newer (Debian 12+, Fedora 36+, recent Arch/Mint also work) |
| CPU | 64-bit Intel/AMD. No ARM build yet |
| API key | free Gemini key at aistudio.google.com/apikey, no card needed, or Claude at console.anthropic.com for the best notes |
First run
Click Settings, choose your provider, paste your key. Your first recording downloads the Whisper model (about 500 MB) and your first PDF fetches the LaTeX packages it needs, so keep internet on for that first run. Everything is cached afterwards and works offline.
Your recordings, transcripts, PDFs and key live in ~/.local/share/notra, outside the app. Replacing the AppImage with a newer version never touches them. Settings shows the folder with a button to open it.
What is in this build
- Single-file Linux app with a native window. No Python, ffmpeg or LaTeX install needed
- Claude or Gemini, switchable in Settings
- Local Whisper transcription with a configurable CPU core count
- Courses tab: your notes grouped course by course
- Notes as real LaTeX: definitions, theorems, derivations, worked examples, questions asked in class, and a ranked Exam Focus list
Known limits
- Linux x86_64 only. macOS and Windows builds are not ready yet
- Needs internet on first run, as above
- The AppImage is 344 MB, most of it the bundled browser engine and transcription runtime