Fisper is a macOS productivity tool that enables system-wide voice dictation by capturing speech and converting it to text directly on the device.

Designed for users who want fast, private, and offline speech-to-text capabilities, Fisper works in any application where you can type — from browsers and editors to terminals and chat apps.

It solves the limitations of built-in dictation and cloud-based services by running AI speech models locally on Apple Silicon, ensuring privacy, speed, and reliability without requiring an internet connection.

Key features include:

• System-wide dictation that works in any macOS app by injecting text at the cursor position.
• Real-time streaming transcription with minimal visual disruption.
• Fully local processing on Apple Silicon (no cloud, no accounts, no telemetry).
• Multiple speech models for speed or accuracy trade-offs.
• Configurable global shortcuts and menu-bar-resident interface.
• Terminal auto-submit mode for hands-free command entry.

Common use cases include hands-free writing, accessibility support, rapid note-taking, coding with voice, and reducing typing workload.

Fisper is distributed as a free-forever macOS app requiring macOS 14 or later on Apple Silicon hardware.