Superwhisper

Superwhisper is a cross-platform AI voice-to-text tool for macOS, Windows, and iOS that turns spoken words into polished, formatted text. Instead of relying on clunky OS-level dictation, it lets you press a global hotkey (⌥+space) anywhere—Slack, Gmail, Notion, Cursor—and start speaking. It’s built for professionals who want to write faster without typing: developers, executives, writers, and anyone dealing with RSI. The core appeal is its adaptability. Predefined modes optimize tone, structure, and formatting for emails, messages, and meetings, while custom modes give you precision control—set formatting rules, structure preferences, and even specialized prompts. You can pick the right model for the job, from top-tier choices like GPT-5, Claude Haiku 4.5, Llama 4, Grok 4.1, and Gemini 3.0 Flash, to on-device options like Whisper Large, S1-mini, and Cohere Transcribe. Recent updates added the S1 family of models, automatic language detection for Cohere Transcribe, and made Cohere Transcribe generally available. Superwhisper works offline with local models, so you can transcribe anytime without Wi-Fi. It supports 100+ languages and dialects, and can translate them all to English. Vocabulary management remembers names, abbreviations, and specialized terms, so you don't have to correct them twice. Clipboard integration auto-pastes your edited transcript into the app you're using. For meetings, the meeting assistant records and digests automatically; for files, you can upload video and audio and get reliable transcripts. Push-to-talk mode gives you hold-to-speak control, and custom shortcuts let you launch, dictate, and control with ease. For developers, Superwhisper integrates with coding agents like Cursor, Claude Code, Open Code, Amp, and Codex, letting you drive an entire fleet of agents faster than you can type. With CLI and MCP server support, it's a hands-free way to pipe context into your workflows. Compared to native dictation or Dragon, Superwhisper's AI