Open Source Speech-to-text AI tools

Voiceink is a macOS dictation application that offers accurate offline voice-to-text transcription. It features customizable dictionaries, automated formatting, and seamless integration for composing emails and messages quickly, enhancing productivity for professionals and students.

opentypeless.com is a voice input app that types your speech wherever your cursor is, with real-time AI polishing for grammar and punctuation. It supports streaming dictation in 99+ languages, customizable hotkeys, and integrates with providers like Whisper and Gemini.

speaktype is a macOS app offering on-device, real-time Apple Silicon–optimized speech-to-text. It keeps audio and transcripts locally, integrates across apps via a keyboard shortcut, supports long-form dictation and contextual prompts, and is open-source.

The TTS Voice Wizard is an AI tool that allows users to convert speech-to-text and back to speech using various speech recognition and text-to-speech methods, and control avatar parameters with voice commands in VRChat.

Whisper Service is a GUI API for OpenAI's Whisper that transcribes audio and video files, allows for corrections, offers customizable options, sends notifications via webhook response, and is deployed on Github under the Apache-2.0 license.

FluidVoice is an open-source macOS voice-to-text dictation app running local speech models and an optional on-device post-processor (Fluid-1) for context-aware formatting and capitalization. It offers system-wide hotkey input, multi-engine, multilingual transcription, and local-first privacy.