Docs
Short pages on setting up embral, getting things done with it, and how it works underneath.
Getting started
- Get started
Install embral, run the setup wizard, and record a first test.
Guides
- Record a meeting
Start, take notes, star moments, and stop.
- Turn on automatic recording
Have embral start when a call starts and stop when it ends.
- Dictate into another app
One hotkey turns speech into cleaned-up text, pasted where your cursor is.
- Name speakers in a transcript
Turn Speaker 1 into a person, during the meeting or after.
- Connect Claude over MCP
Give Claude and other AI assistants read access to your meetings and dictations.
- Export notes to Obsidian
Save a Markdown copy of every finished meeting into your vault, or any folder.
- Turn on cloud processing
Sign in, get hours, and choose what happens when they run out.
- Import a recording
Turn an audio file you already have into a meeting.
- Change transcription accuracy or language
Pick between speed and accuracy, or switch to the multilingual model.
Concepts
- How recording works
No bot joins your call. embral records what your computer plays and hears.
- Local and cloud processing
Everything runs on your machine by default. The cloud tier is optional, per feature, and stores nothing.
- Where your data lives
One folder on your disk, in formats any tool can read.
- How speakers get their names
embral separates voices, but names come from you or your notes, never from voice recognition.