Whisper

approved

by nikdanilov

Speech-to-text using OpenAI Whisper.

352 stars46,011 downloadsUpdated 1mo agoMIT

Whisper — Speech-to-text for Obsidian

Record or upload audio, transcribe with Whisper, and optionally post-process the result with an LLM. Works on desktop and mobile.

Works with OpenAI, Groq, Azure, or any other Whisper-compatible API.

Quick Start

  1. Install from Settings → Community Plugins → search "Whisper"
  2. Add your API key in the plugin settings
  3. Open a note, press Alt + Q, speak, press Alt + Q again

The transcription appears at your cursor.

Usage

Record — click the mic icon in the sidebar, or press Alt + Q to start/stop.

Upload — command palette → Upload audio file (mp3, mp4, m4a, wav, webm, ogg).

Right-click — right-click any audio file in your vault → Transcribe audio file.

All commands can be assigned custom hotkeys in Obsidian's hotkey settings:

  • Start/stop recording (Alt + Q by default)
  • Pause/resume recording
  • Open recording controls
  • Upload audio file

Automation

Trigger from iOS Shortcuts, Alfred, or any tool that can open URLs:

obsidian://whisper                 open controls
obsidian://whisper?command=start   start recording
obsidian://whisper?command=stop    stop and transcribe
obsidian://whisper?command=pause   pause/resume
obsidian://whisper?command=cancel  discard recording

Post-Processing

Enable Post-processing in settings to run transcriptions through an LLM — fix grammar, remove filler words, format as markdown, extract action items.

Supports Claude, GPT, or any OpenAI-compatible endpoint (Ollama, LM Studio, etc.).

You can also enable Auto-generate title to create descriptive filenames for your notes.

Note Templates

When Create note file is enabled, you can customize the filename and content using template variables:

VariableExample
{{title}}Meeting Notes
{{audioFile}}recordings/2026-04-05.webm
{{transcription}}the transcribed text
{{date}}2026-04-05
{{time}}14-30-00
{{datetime}}2026-04-05 14:30:00

Example note template:

# {{title}}
![[{{audioFile}}]]

{{transcription}}

Use ![[{{audioFile}}]] to embed audio (playable) or [[{{audioFile}}]] to link.

Manual Installation

Download manifest.json, main.js, styles.css from releases into .obsidian/plugins/whisper/ in your vault.

Contributing

Issues and PRs welcome — GitHub Issues


Buy me a coffee · @nikdanilov_

For plugin developers

Search results and similarity scores are powered by semantic analysis of your plugin's README. If your plugin isn't appearing for searches you'd expect, try updating your README to clearly describe your plugin's purpose, features, and use cases.