Meeting Scribe

Record meetings, transcribe with cloud STT, summarize with LLM, and auto-generate structured meeting notes.

Features

• One-click recording — Start/stop with a ribbon icon, status bar, or command palette
• Cloud STT transcription — OpenAI Whisper, CLOVA Speech, or Gemini
• LLM summarization — OpenAI or Anthropic Claude generate structured summaries
• Transcript sidebar — Interactive chat-bubble view with audio playback sync
• Inline editing — Edit transcription text, timestamps, and speaker names directly
• Speaker mapping — Rename speakers to real names with vault note autocomplete and wiki-links
• Re-summarize — Regenerate summaries after editing the transcript
• Export — Export edited transcripts to Markdown
• Import audio — Process existing audio files from your vault
• Recording formats — WebM, M4A, or WAV (configurable)
• BYOK — Bring your own API keys. No intermediary servers, no subscriptions

Screenshots

Session list & Transcript view

Sessions	Transcript

Editing & Speaker mapping

Inline editing	Speaker mapping

Audio player & Re-summarize

Volume control	Re-summarize

Settings

How it works

1. Press the ribbon icon or use the command palette to start recording
2. Stop recording when your meeting ends
3. The plugin processes your audio in the background:
   • Transcribes speech to text via cloud STT
   • Summarizes the transcript with an LLM
   • Generates a structured meeting note in your vault
4. The transcript sidebar opens automatically with an interactive view
5. Edit transcription errors, rename speakers, and re-summarize as needed

Setup

1. Install the plugin from Obsidian community plugins
2. Open Settings and enter your API keys:
   • STT: OpenAI, CLOVA Speech, or Gemini
   • LLM: OpenAI or Anthropic
3. Click Run test to verify your setup
4. Start recording

Settings

Setting	Description	Default
STT provider	Speech-to-text service (OpenAI, CLOVA Speech, Gemini)	OpenAI
STT model	Transcription model	gpt-4o-mini-transcribe
STT language	Transcription language	Auto-detect
LLM provider	Summarization service (OpenAI, Anthropic)	Anthropic
Summary language	Language for generated summaries	Auto-detect
Notes folder	Where meeting notes are saved	Meeting Notes
Audio folder	Where recordings are stored	_attachments/audio
Recording format	Audio format for recordings (WebM, M4A, WAV)	WebM
Audio retention	Keep or delete audio after processing	Keep
Include transcript	Append full transcript in generated notes	On
Separate transcript file	Create a separate transcript Markdown file	Off
Enable summary	Generate LLM summary (disable for transcript-only)	On
Auto-open sidebar	Open transcript sidebar when a meeting note is active	On
Consent reminder	Show recording consent notice before starting	On

Commands

Command	Description
Start recording	Begin audio recording
Stop recording	Stop recording and start processing
Toggle recording	Start or stop recording
Import audio file	Process an existing audio file
Open transcript sidebar	Open the interactive transcript view
Audio play/pause	Play or pause audio playback
Audio skip back	Skip back 5 seconds
Audio skip forward	Skip forward 5 seconds

Transcript sidebar

The sidebar provides an interactive workspace for your transcripts:

• Chat bubbles — Each speaker segment displayed as a colored bubble with timestamp
• Click to seek — Click any timestamp to jump to that point in the audio
• Inline editing — Click bubble text to edit transcription errors
• Delete/Split — Hover to reveal delete and split buttons (split available in edit mode)
• Speaker mapping — Click a speaker name to rename with vault autocomplete
• Wiki-links — Link speaker names to existing vault notes
• Re-summarize — Regenerate the summary after edits (uses LLM API)
• Export — Export the transcript as a Markdown file
• Playback sync — Current segment highlights during audio playback
• Keyboard shortcuts — Configurable hotkeys for play/pause, skip forward/back

Requirements

• Obsidian v0.15.0+
• API key for at least one STT provider
• API key for at least one LLM provider (optional if summary disabled)
• Microphone access (for recording; not needed for audio import)

Privacy

This plugin uses your own API keys to connect directly to provider APIs. No data passes through intermediary servers. Audio and transcripts are sent only to the provider you configure. Your vault data stays local.

License

MIT