crosswk/SayIt

Open-source voice typing for Windows — a Wispr Flow / Superwhisper alternative. Press a shortcut, speak, and AI-polished text lands at your cursor. Local models, your own API keys, or a self-hosted backend.

What it solves

SayIt is a voice typing tool for Windows that allows users to dictate text into any application. It addresses the slow speed of manual typing and the inaccuracy of raw speech-to-text by providing AI-powered cleanup to remove filler words and repair recognition errors, ensuring the polished text is inserted directly at the cursor.

How it works

Users trigger a shortcut to record audio, which is then processed through a chosen voice engine. The system supports three operational modes: Local mode (using GGUF models on the user's GPU), Cloud API mode (connecting directly to providers like Groq or Doubao), and Server mode (connecting to a self-hosted FastAPI backend). Once transcribed, the text can be passed through an AI cleanup phase using OpenAI-compatible endpoints to refine the grammar and tone before being typed into the active window.

Who it’s for

It is designed for Windows users who want to increase their writing speed and productivity across various software, including editors, chat apps, and browsers, while maintaining control over how their speech is processed and privacy settings.

Highlights

  • Voice typing anywhere: Inserts text directly into any Windows application without window switching.
  • Editable AI cleanup: Customizable prompts to remove filler words or format ideas.
  • Context-aware writing: Can read surrounding text to match tone or act as an editing instruction for selected text.
  • Flexible recognition: Supports local GGUF models, cloud APIs, and self-hosted servers.
  • Overlay feedback: Provides a visual waveform and optional live captions during recording.
  • Per-app rules: Automatically switches cleanup behavior based on the active application.

Related

  • Project
    OpenWhispr/openwhispr

    OpenWhispr is a cross‑platform desktop app that converts speech to text, provides live meeting transcription with speaker diarisation, and lets you talk to AI assistants (GPT‑5, Claude, Gemini, etc.) directly from your keyboard. It works offline with Whisper‑based models or via cloud services, stores notes locally (with optional cloud sync), and offers an API for programmatic use.

  • Project
    tover0314-w/opentypeless

    An open-source AI voice input tool for macOS, Windows, and Linux that provides context-aware dictation, AI rewriting, and one-shot voice Q&A across any desktop application.

  • Project
    theJayTea/WritingTools

    Writing Tools is a free, open‑source desktop app for Windows, Linux and macOS that lets you invoke an LLM anywhere on your computer with a hot‑key. It can proofread, rewrite, change tone, translate, summarise web pages or YouTube transcripts, and even chat with the model. You can use cloud APIs (Gemini, OpenAI, Anthropic, etc.) or run local models via Ollama, llama.cpp or Apple‑silicon MLX. The app is privacy‑focused (no telemetry, local storage of API keys) and licensed under GPL‑v3.

  • Project
    Quantatirsk/qwen3-asr

    Qwen3‑ASR is a self‑hosted speech‑to‑text server that wraps Qwen3‑ASR models (0.6 B & 1.7 B). It auto‑selects a GPU‑vLLM backend or a CPU‑Rust backend, offers OpenAI‑ and Alibaba‑compatible HTTP/WebSocket APIs, supports speaker diarization, VAD‑based segmentation, batch inference, and can be deployed via Docker, custom GPU images, or offline tarballs. MIT‑licensed.

  • Project
    mkiol/dsnote

    Speech Note is a Linux/Sailfish desktop app for taking, reading and translating notes entirely offline. It combines multiple open‑source STT engines (Coqui STT, Vosk, Whisper cpp, Faster Whisper, April‑ASR), many TTS voices (eSpeak‑NG, Piper, RHVoice, Mimic 3, etc.) and the Bergamot offline translator. Users download model checkpoints through an integrated browser, keeping all processing on‑device for privacy. Available as a Flathub Flatpak and also buildable from source, it is GPL‑3.0 licensed.