Superwhisper — AI Voice-to-Text for Mac, Windows & iOS

Superwhisper is an AI-powered voice dictation app that turns speech into polished text across any app. It supports offline transcription, 100+ languages, customizable AI modes, and integrates with coding tools like Cursor and Claude Code.

What is Superwhisper?

Superwhisper is a voice-to-text application that uses OpenAI's Whisper models (and other LLMs) to transcribe speech into formatted, context-aware text. Unlike basic dictation tools, it goes beyond raw transcription by applying AI post-processing to clean up grammar, adjust tone, and format output based on configurable modes.

It works system-wide — activate it with ⌥ + Space, speak, and the transcribed text is automatically pasted into whatever app you're using (Slack, Gmail, Cursor, Notion, etc.).

Key Features

Offline mode — Runs locally on-device using Whisper models, so audio never leaves your machine. No Wi-Fi required.

Predefined & custom modes — Modes like "Formal", "Casual", "Legal", and "Chat" adjust tone, structure, and formatting. You can also build fully custom modes with your own system prompts.

Model flexibility — Choose from multiple AI providers for post-processing: GPT, Claude, Llama, Grok, Gemini, Mistral. You can also bring your own API keys.

100+ languages — Supports multilingual transcription and can translate any language to English.

Clipboard integration — Transcribed text is auto-pasted into the active app. No copy-paste needed.

Meeting assistant — Record and transcribe meetings with automatic note generation.

File transcription — Upload audio/video files for transcription.

Push to talk — Hold, speak, release. Simple control flow.

Custom vocabulary — Add names, abbreviations, and domain-specific terms for better accuracy.

Agentic coding workflows — Works with Cursor, Claude Code, Open Code, Amp, and other coding agents for hands-free development.

Platform Availability

Platform Status
macOS Full support (Apple Silicon + Intel)
Windows Supported
iOS Available on App Store
Linux Not supported

Intel Macs work best with cloud models; offline/local models perform best on Apple Silicon.

Pricing

Plan Price Includes
Free $0 Voice-to-text in any app, meeting recording, 100+ languages, small AI models, custom prompts
Pro $8.49/mo (40% off for students) Cloud + local AI models, own API keys, translation, file transcription, priority support
Yearly ~$85/year (2 months free) Same as Pro
Lifetime $249 One-time payment, all Pro features
Enterprise Custom SOC 2 Type II, centralized billing, model access control

Pro features include a 15-minute free trial. 30-day refund policy on all paid plans.

Notable Users & Endorsements

Superwhisper has gained traction among developers and tech founders. Andrej Karpathy popularized the concept of "vibe coding" — using Superwhisper with Cursor to build software almost entirely by voice. Guillermo Rauch (Vercel CEO) and Pieter Levels have also publicly endorsed the tool.

Companies like Vercel, Shopify, OpenAI, Meta, and Spotify are listed as users on the website.

Competitors

Tool Price Offline Platform
Wispr Flow $15/mo No Mac
Sotto $29 one-time Yes Mac
Voibe $99 lifetime Yes Mac
VoiceInk $39 one-time Yes Mac
BetterDictation $24 one-time Yes Mac
OpenWhispr Free (OSS) Yes Mac, Win, Linux
Speakmac $19 one-time Yes Mac
macOS Dictation Free Yes Mac

Superwhisper differentiates through its depth of customization (custom modes, model selection, app-specific settings) and cross-platform availability, but comes at a higher price point than most alternatives.

Links


Superwhisper logo


Superwhisper docs


Superwhisper roadmap

links

social