tover0314-w/opentypeless

What it solves

OpenTypeless is an open-source AI voice input tool for macOS, Windows, and Linux that goes beyond simple transcription. It solves the problem of tedious typing and the lack of context-aware AI rewriting in standard dictation tools, allowing users to speak naturally and have the AI polish the text based on the active application (e.g., email, chat, or code) before inserting it into the app.

How it works

Users trigger the tool via global hotkeys to capture audio, which is then processed through a chosen Speech-to-Text (STT) provider. The resulting raw transcript is optionally sent to a Large Language Model (LLM) for "AI polish"—where the tool detects the active application locally to adapt the tone and structure. The final text is then simulated as keyboard input or copied to the clipboard.

Who it’s for

It is designed for power users who want to speed up their writing workflow across any desktop application, as well as those who prefer open-source control over their STT and LLM providers (BYOK - Bring Your Own Key).

Highlights

• App-aware writing: Automatically detects the active app to adapt writing style and tone.
• Ask Anything: A one-shot voice Q&A flow that provides concise answers in a floating note without opening a chat app.
• Selected-text editing: Allows users to rewrite, summarize, or translate text already selected in another application via voice instructions.
• Provider flexibility: Supports a wide range of STT (e.g., Groq, Deepgram, OpenAI Whisper) and LLM (e.g., Claude, Gemini, DeepSeek, Ollama) providers.
• Custom dictionary: Includes local correction rules and custom terms to fix recurring transcription mistakes.