Otter.ai
AI meeting assistant for transcription and automated note-taking
OpenWhispr
Dictate anywhere, get instant AI-powered transcription with privacy options
OpenWhispr lets you dictate text via a global hotkey, transcribing speech locally or via cloud AI models, with customizable agents, history, and cross‑platform support.
Overview
Overview
OpenWhispr is a desktop dictation tool that lets users capture speech and turn it into text with a single customizable hotkey. Designed for professionals, developers, and anyone who prefers hands‑free typing, it works on macOS, Windows, and Linux and integrates seamlessly with any active window.
Features and Deployment
The app offers both local Whisper transcription and cloud‑based AI providers such as OpenAI, Anthropic Claude, and Google Gemini, giving you the choice between maximum privacy and fastest turnaround. Local models can be downloaded in sizes ranging from tiny to large‑turbo, and a built‑in cleanup removes cached files with one click. A modern React‑based UI provides a draggable panel, automatic pasting at the cursor, and a history stored in a local SQLite database. Configuration—including API keys, hotkey mapping, and agent naming—is handled through an in‑app control panel. Installation requires Node 18+ (and optional Python for local models); the app can be run in development mode or packaged as a standalone executable for personal use.
Highlights
Customizable global hotkey for instant dictation start/stop
Choice of local Whisper or cloud AI providers (OpenAI, Anthropic, Google)
Privacy‑first local processing with model download and one‑click cleanup
Modern draggable UI with automatic pasting and SQLite transcription history
Pros
- Flexible processing options for speed or privacy
- Supports latest AI models like GPT‑5, Claude 4.1, Gemini 2.5
- Cross‑platform compatibility (macOS, Windows, Linux)
- Simple setup with hotkey and automatic pasting
Considerations
- Requires Node.js 18+ and optional Python for local models
- Large Whisper models can consume significant disk space
- Cloud processing needs API keys and an internet connection
- macOS globe‑key feature requires Xcode command‑line tools
Managed products teams compare with
When teams consider OpenWhispr, these hosted platforms usually appear on the same shortlist.
Looking for a hosted option? These are the services engineering teams benchmark against before choosing open source.
Fit guide
Great for
- Remote workers needing hands‑free note taking
- Developers who want AI assistance without leaving the editor
- Privacy‑conscious users preferring local transcription
- Users on multiple operating systems seeking a consistent dictation tool
Not ideal when
- Environments without microphone or required permissions
- Users unwilling to download large Whisper models
- Teams that require built‑in collaboration or sharing features
- Scenarios needing real‑time captioning for live events
How teams use it
Write emails while coding
Dictate email drafts, have the AI polish language, and paste directly into the email client without switching apps.
Create meeting notes on the fly
Start the hotkey during a meeting, speak, and have transcribed, timestamped notes saved in history for later export.
Generate code snippets with AI assistance
Dictate a description, let the selected AI provider generate code, and insert it instantly into the IDE.
Maintain privacy‑first transcription
Download local Whisper models and transcribe sensitive conversations without sending audio to external services.
Tech snapshot
TypeScript59%
JavaScript36%
Python3%
CSS1%
Shell1%
Swift1%
Frequently asked questions
Do I need an internet connection?
Local Whisper works offline; cloud providers need internet and API keys.
Which operating systems are supported?
macOS 10.15+, Windows 10+, and major Linux distributions.
How is my voice data handled?
With local processing, audio never leaves your machine; cloud processing sends audio to the selected provider per their policies.
Can I change the hotkey?
Yes, the hotkey is configurable in the Settings panel.
How do I manage Whisper model size?
The app lets you download specific model variants (tiny to large) and provides a one‑click cleanup to remove cached models.
Project at a glance
Active- Stars
- 807
- Watchers
- 807
- Forks
- 99
LicenseMIT
Repo age7 months old
Last commit15 hours ago
Primary languageTypeScript
Last synced 12 hours ago