FluidVoice for MAC
Free, open-source voice-to-text dictation for Mac: speak into any text field, with local speech models, optional AI cleanup and meeting transcripts.
FluidVoice Overview
FluidVoice is a dictation app for macOS that turns your speech into text in any app. You press a shortcut, talk, and the words appear in the text field you are using, whether that is an email, a document or a chat box. Speech recognition runs on your Mac with local models, so dictation works offline, and the app is optimized for Apple Silicon. Different modes, such as Write, Command and Dictate, change what happens with what you say.
For cleaner text, FluidVoice can enhance the transcript after you speak. You can connect your own AI provider, such as OpenAI or Groq, or use Fluid Intelligence, a fully local model for formatting that the developers keep separate from the open-source code. FluidMeet records and transcribes meetings with speaker names and can summarize them on your Mac or with your own AI provider. FluidVoice is made by Altic, is free and open source under the GPL-3.0 license, and needs macOS 15 or later. The official website is altic.dev/fluid; the developers warn that fluidvoice.org is not theirs.
Key Features
FluidVoice lets you write by speaking.
- System-wide dictation: speak into any text field in any app.
- Real-time: words appear as you speak.
- Offline: local speech models run on your Mac.
- Modes: Write, Command and Dictate for different jobs.
- AI cleanup: on-device Fluid Intelligence or your own AI provider.
- Languages: 40+ languages across the available models.
- FluidMeet: meeting transcripts with speakers and summaries.
- Apple Silicon: optimized with Core ML and Metal.
What’s New in Version 1.6.12
FluidVoice 1.6.12 improves FluidMeet meeting transcripts.
- Your own start/stop meeting shortcut in Shortcuts settings
- Export transcripts as Markdown, text or JSON with timestamps and speaker names
- Clearer microphone speech when other audio is playing
- Edit transcript text in place, rename and merge speakers, and undo corrections
- Meeting history loads details as needed
- Turn FluidMeet on or off without deleting saved meetings
FluidVoice FAQ
Is FluidVoice free?
Yes. FluidVoice is free and open source under the GPL-3.0 license.
Does FluidVoice work offline?
Yes. Speech recognition runs on your Mac with local speech models, so dictation works without an internet connection.
Can I dictate into any app?
Yes. FluidVoice types your speech into any text field in any app.
What is Fluid Intelligence?
It is a fully local AI model for on-device dictation cleanup, such as formatting. It is maintained privately, separate from the open-source FluidVoice code.
Can I use my own AI provider?
Yes. Optional AI post-processing can use providers such as OpenAI, Groq or a custom provider with your own key.
Can FluidVoice transcribe meetings?
Yes. FluidMeet records and transcribes meetings with speaker names, exports transcripts and can summarize them on your Mac or with your own AI provider.
Which macOS version does FluidVoice need?
macOS 15 or later.
Is fluidvoice.org the official website?
No. The developers say fluidvoice.org is not affiliated with them. The official site is altic.dev/fluid, and official downloads are on GitHub.
Can I install FluidVoice with Homebrew?
Yes. Run brew install --cask fluidvoice.
Where does the FluidVoice download on fluidvoice.FULLMAC.app come from?
The download button leads to the official FluidVoice releases on GitHub, https://github.com/altic-dev/FluidVoice/releases/latest. fluidvoice.FULLMAC.app does not change or repackage the app.
Similar Apps
Live TranslatorLive Translator is a free, open-source Mac app that translates whatever audio your Mac is playing, such as YouTube videos, podcasts, meetings and movies, and shows the translation live in a floating panel. It sits in the menu bar, captures system audio through Apple's ScreenCaptureKit, turns the speech into text and sends the text to an AI model that writes a natural translation. The AI sees the whole conversation, so the translation reads like a growing document rather than word-by-word subtitles, and new parts are highlighted.AI
JanJan is a free, open-source app for running AI models on your own Mac. You can download large language models such as Llama, Gemma, Qwen and GPT-oss from Hugging Face and chat with them locally, so everything can stay on your computer. When you want, Jan can also connect to cloud models from OpenAI, Anthropic, Mistral, Groq, MiniMax and others with your own API keys. Custom assistants, a local OpenAI-compatible API server on localhost:1337 and Model Context Protocol support make it useful for agent and developer work too.AI
ClaudeBarClaudeBar is a menu bar app that shows how much of your AI coding quota is left. It reads the limits of the coding assistants you use, such as Claude, Codex, Gemini, Copilot and Cursor, and many more providers including OpenRouter, DeepSeek, Mistral and Windsurf. For each one it shows what is used and what remains, counts down to the next reset and sends a notification before you run out, so you are not stopped in the middle of a task.AI
AionUiAionUi is a desktop app for working with AI agents. It ships with its own agent engine, so it works as soon as it is installed: you paste an API key from Gemini, OpenAI, Anthropic or another of more than 30 supported platforms, or point it at a local model in Ollama or LM Studio. The agent can read and write files, search the web, generate images and use MCP (Model Context Protocol) tools, and 21 ready-made assistants help with tasks such as creating PowerPoint, Word and Excel files.AI
VoiceInkVoiceInk turns your voice into text on the Mac. You press a shortcut, speak, and the transcript is typed into whatever app you are using. Transcription runs locally on your Mac, and optional cloud enhancement can clean up and restructure what you said. Custom words teach it names and terms, Smart Replace expands short phrases you define, and it removes filler words and repetition so your text reads cleanly.AI
SuperwhisperSuperwhisper is a voice-to-text app for the Mac. You press a shortcut, speak, and your words are typed into whatever app you are using, from email and documents to code editors and chat windows. It understands more than 100 languages, can run Whisper speech models locally on your Mac so dictation works offline, and lets you add your own vocabulary so names and technical terms come out right. Modes such as Voice, Message and Email shape the text for the job, and custom prompts let you build your own.AI
FULLMAC.app does not host this app. Download buttons lead to the developer’s own pages.