Ito is an open-source AI-powered voice dictation tool for macOS that transforms your spoken words into smart, contextually appropriate content. It activates with a customizable hotkey in any text field across applications like Slack, Notion, Google Docs, and Cursor.
Standard tools like Apple’s built-in dictation or Google Docs Voice Typing often work only within specific apps or browsers. Ito is designed to work everywhere, offering a single, consistent experience across your entire Mac.
Features
- Across Applications: Works across any macOS application with text input fields, including productivity apps, code editors, and communication platforms.
- Intent-Based Processing: Transforms casual spoken instructions into properly formatted content, from full emails to code snippets.
- Customizable Trigger Keys: Configure any keyboard shortcut combination to activate the voice input system-wide.
- Real-Time Transcription: High-accuracy speech-to-text processing powered by advanced AI models with instant text insertion.
- Custom Vocabulary Support: Add technical terms, names, and specialized language to improve recognition accuracy.
- Smart Content Formatting: Automatically structures output based on context, whether you’re writing emails, creating documentation, or coding.
- Open-Source Architecture: Full transparency with a customizable codebase available on GitHub.
- Local Processing Options: Voice data can be processed locally for enhanced privacy control.
- Notes and History System: Automatically saves transcriptions for later reference and tracks usage patterns.
Use Cases
- Email Composition: Speak natural instructions like “Tell Sarah thanks and reschedule for next week” to generate complete, professionally formatted emails.
- Code Development: Voice commands such as “Make a dark mode toggle in React” produce functional code snippets ready for implementation.
- Document Creation: Convert meeting notes into structured product briefs by speaking summary instructions.
- Team Communication: Generate polished Slack messages by stating your intent, like “Ask James if we can move the meeting to 3pm”.
- Meeting Management: Create organized agendas with bullet points by speaking simple instructions about upcoming meetings.
How to Use It
1. Get the latest version from the official website or GitHub repo and drag the application into your Applications folder.
2. When you first run Ito, it will ask for two key permissions. It needs Microphone access to hear your voice and Accessibility access to listen for your global keyboard shortcut and type text into other apps.
3. Complete the initial authentication process using Google, Apple, GitHub through Auth0, or create a local account. The guided onboarding process will walk through basic configuration options and test your microphone setup.
4. Configure your preferred trigger key combination (the default is Fn + Space). Test the microphone during setup to ensure clear audio input.
5. To use Ito, hold your configured trigger key and speak naturally into any text field. The application will process your speech and automatically insert the formatted text.
6. Access settings to adjust voice sensitivity levels, modify keyboard shortcuts, and add custom vocabulary terms. The application learns from usage patterns over time, improving accuracy for your specific voice and common phrases.
Pros
- Truly Universal: It works in almost any text field you can click on.
- High Accuracy: The transcription quality is excellent, even handling specialized vocabulary well once you’ve added it to the custom dictionary.
- Open Source and Privacy-Focused: Being open source means its code is transparent and can be audited. The option for local data processing is a huge plus for privacy-conscious users.
- Free to Use: The core functionality is completely free.
Cons
- macOS Only: Currently, Ito is only available for Mac users.
- Requires System Permissions: Granting Accessibility access is required for Ito to work. While necessary for its function, some users may be hesitant to provide this level of permission.
- Internet-Reliant for Best Performance: While local processing is an option, the highest accuracy models often rely on a cloud-based server, which means it works best with an active internet connection.
FAQs
Q: What makes Ito different from built-in macOS dictation?
A: Ito uses large language models to understand your intent and generate contextually appropriate content, while built-in dictation simply converts speech to text without formatting or interpretation. For example, saying “email John about the meeting” would produce a complete email draft in Ito, while standard dictation would just type those exact words.
Q: How accurate is the voice recognition compared to other dictation tools?
A: Ito focuses on intent recognition rather than perfect word-for-word transcription, which can actually be more useful for productivity tasks. The accuracy improves over time as the system learns your voice patterns and common phrases. Custom vocabulary support helps with technical terms and proper names that standard dictation might miss.
Q: Is my voice data secure and private?
A: Ito provides local processing options that keep voice data on your device, with encrypted transmission for any cloud features. As an open-source project, you can audit the source code to understand exactly how your data is handled and stored.
Q: What are the system requirements for running Ito?
A: The application requires macOS 10.15 or later, with microphone access and accessibility permissions. For development or advanced customization, you’ll need Node.js 20+, Bun, and the Rust toolchain, though regular users only need to download and install the pre-built application.
