Getting started

What you need

NeuroBridgeEDU runs on most laptops less than five years old. The minimums are forgiving; the recommendations are where the experience starts to feel instant.

• OS: macOS 13+ on Apple Silicon, Windows 10 64-bit, or Ubuntu 20.04+. Latest stable in each family is best.
• RAM: 8 GB minimum, 16 GB recommended (lets you run the larger summarisation models without paging).
• Disk: ~4 GB free for the lightweight setup, ~10 GB if you want the larger models on disk.
• Microphone: any built-in mic works; a USB headset is better for long lectures.
• GPU: not required, but if you have Apple Silicon, an NVIDIA card, or a Vulkan-capable AMD/Intel GPU, transcription gets three to ten times faster.
• Internet: needed for model downloads and Pro cloud. Recording, Whisper transcription, and Built-in AI summaries all work offline on every platform after download. Ollama and optional Pro cloud are additional provider choices.

System audio capture

If you also want to record what's playing on your computer (e.g. a Zoom call):

• macOS — needs a virtual audio device (BlackHole 2ch is free).
• Windows — built in via WASAPI, no extra setup.
• Linux — needs a PulseAudio or PipeWire loopback source.

1. Install the app

NeuroBridgeEDU is not publicly available yet. The free app is coming soon; this guide covers how installation will work at launch.

At launch, installers are planned for all three platforms:

• macOS: an Apple Silicon .dmg you open and drag into Applications.
• Windows: a .msi or .exe installer.
• Linux: .deb, .rpm, or AppImage packages, whichever fits your distro.

2. Walk through onboarding

First launch takes you through a short setup with seven chapters. The whole thing is keyboard-navigable.

1. Welcome — one-screen intro with the project values.
2. Language — pick your UI language (changeable later in Settings).
3. Models — download the default Whisper model and a small built-in summarisation model. The only step that needs internet, and you only do it once.
4. Reading style — pick how detailed you want summaries to be by default.
5. Permissions: macOS only, grant Microphone. System audio uses a virtual audio device such as BlackHole in v2.0.
6. Telemetry — opt in or out of anonymous error reporting. Off is fine.
7. Complete — done.

There's no in-app "Reset onboarding" button right now. The only way to redo onboarding is to Delete All Data in Settings → Storage, which also wipes your meetings — so use it deliberately.

3. Make your first recording

You should land on the Home screen.

1. Pick your microphone — open the device dropdown in the recording controls and select your input by name. Avoid "Default".
2. (Optional) Pick a system audio source — if you want to capture audio playing on your computer.
3. Start recording — press Cmd+Shift+R (macOS) or Ctrl+Shift+R (Windows / Linux), or click the red Record button.
4. Speak — the live transcript scrolls in as you talk. The audio meter should move with your voice.
5. Stop — same shortcut again, or click Stop, or pressEscape. (Space pauses and resumes mid-recording but doesn't start one.)

Your recording is transcribed, then the audio file is deleted automatically for privacy and the transcript is saved. The meeting appears in your list. To get a summary, open the meeting, choose a template, and press Generate.

4. Open your meeting

Click Meetings in the sidebar. Your new session sits at the top of the list. Inside the meeting view you get:

• Summary - open the meeting, pick a template, and press Generate to create the AI notes.
• Transcript — the full text, directly editable.
• Action Items — Tasks, Decisions, and Open Questions (Pro).
• Chat — open Study Sage with Cmd+Shift+K / Ctrl+Shift+K and ask questions about the session.
• Recall Engine — generate flashcards from the summary (Pro).
• Export — save the summary as a PDF (Pro).

To create a summary, choose the template above the summary and click Generate. To change the format later, choose another template and click Regenerate.

5. Adjust your preferences

Open Settings from the sidebar. The most useful first changes:

• Transcription language — force a single language for better accuracy.
• Whisper model — trade speed for accuracy.
• Summary provider: Built-in AI (on-device on macOS, Windows, and Linux) or Ollama on Free; Pro cloud keys for Mistral, OpenAI, or Anthropic.
• Theme — Light, Dark, or System (also via the sidebar toggle).
• UI language: 32 supported.

Keyboard shortcuts worth knowing

• Cmd+Shift+R / Ctrl+Shift+R — start or stop recording.
• Space — pause / resume during recording.
• Escape — stop recording.
• Cmd+K / Ctrl+K — open the Command Palette.
• Cmd+Shift+K / Ctrl+Shift+K — open Study Sage.
• Cmd+Shift+F / Ctrl+Shift+F — toggle focus mode.
• Tab / Shift+Tab — move between interactive elements.

You're set up

From here, depending on what you want next:

• Read How it works to understand the local-first AI pipeline.
• Read Features for the full surface and when to reach for each capability.
• Read Privacy and data for the technical detail on what stays local and how API keys are protected.
• If something's broken, jump to Troubleshooting.