FluidVoice

Fully open source voice-to-text dictation app for macOS with AI enhancement.

Get the latest release from here

[!IMPORTANT] This project is completely free and open source. If you find FluidVoice useful, please star the repository. It helps with visibility and motivates continued development. Your support means a lot.

Latest Update

• Added Cohere Transcribe support for higher-accuracy multilingual dictation on Apple Silicon
• Expanded the offline voice engine lineup with Parakeet v3/v2, Cohere, Apple Speech, and Whisper
• Continued improving the multilingual dictation experience for users who switch between languages often

Star History

Demo

Command Mode - Take any action on your mac using FluidVoice

https://github.com/user-attachments/assets/ffb47afd-1621-432a-bdca-baa4b8526301

Write Mode - Write/Rewrite text in ANY text box in ANY App on your mac

https://github.com/user-attachments/assets/c57ef6d5-f0a1-4a3f-a121-637533442c24

Screenshots

Command Mode Preview

FluidVoice History

New Features (v1.5)

• Overlay with Notch support
• Command Mode
• Write Mode
• New History stats
• Stats to monitor usage

Features

• Live Preview Mode: Real-time transcription preview in overlay
• Multiple Speech Models: Parakeet TDT v3 & v2, Cohere Transcribe, Apple Speech, and Whisper
• Real-time transcription with extremely low latency
• AI enhancement with OpenAI, Groq, and custom providers
• Global hotkey for instant voice capture
• Smart typing directly into any app
• Menu bar integration for quick access
• Auto-updates with seamless restart
• Opt-in beta channel for early preview builds

Supported Models

Model	Best for	Language support	Download size	Hardware
Parakeet TDT v3	Fast default multilingual dictation	25 languages	~500 MB	Apple Silicon
Parakeet TDT v2	Fastest English-only dictation	English only	~500 MB	Apple Silicon
Cohere Transcribe	High-accuracy multilingual dictation	14 languages	~1.4 GB	Apple Silicon
Apple Speech	Zero-download native macOS speech recognition	System languages	Built-in	Apple Silicon + Intel
Whisper Tiny / Base / Small / Medium / Large	Broad compatibility, including Intel Macs	99 languages	~75 MB to ~2.9 GB	Apple Silicon + Intel

Notes: Parakeet TDT v3 is the default on Apple Silicon. Cohere is the stronger pick if you want a larger multilingual model with higher displayed accuracy. Whisper remains the fallback for Intel Macs and the widest language coverage.

Parakeet TDT v3 Languages

Bulgarian, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Spanish, Swedish, and Ukrainian.

Parakeet TDT v2 Languages

English.

Cohere Transcribe Languages

English, French, German, Italian, Spanish, Portuguese, Greek, Dutch, Polish, Mandarin, Japanese, Korean, Vietnamese, and Arabic.

Apple Speech Languages

System language support depends on the macOS speech recognition languages available on your machine.

Whisper Language Support

Whisper supports up to 99 languages, depending on the model size you choose.

Quick Start

1. Download the latest release
2. Move to Applications folder
3. Grant microphone and accessibility permissions when prompted
4. Set your preferred hotkey in settings
5. Optionally add an AI provider API key for enhanced transcription, keys are stored securely in your macOS Keychain. Make sure select "Always allow" for permissions
6. Optional: opt in to beta builds in Settings → Automatic Updates → Beta Releases

Requirements

• macOS 15.0 (Sequoia) or later
• Apple Silicon Mac (M1, M2, M3, M4)
• Intel Macs are supported from 1.5.1 builds using Whisper models!
• Microphone access
• Accessibility permissions for typing

Join our small community to help us grow and give feedback :) ( Or just hang?!)

https://discord.gg/VUPHaKSvYV

Building from Source

git clone https://github.com/altic-dev/Fluid-oss.git
cd Fluid-oss
open Fluid.xcodeproj

Build and run in Xcode. All dependencies are managed via Swift Package Manager.

Contributing

Contributions are welcome! Please create an issue first to discuss any major changes or feature requests before submitting a pull request.

Setting Up Your Development Environment

1. Clone the repository:

   git clone https://github.com/altic-dev/Fluid-oss.git
   cd Fluid-oss
   

2. Open in Xcode:

   open Fluid.xcodeproj
   

3. Run from Xcode (one-time signing setup):

   • Target: FluidVoice → Signing & Capabilities
   • Enable Automatically manage signing
   • Pick your Team (Personal Team is fine)
   • This is stored in xcuserdata/ (gitignored), so it won’t affect your PR

4. Build and run - All dependencies are managed via Swift Package Manager

5. Build only (no signing):

   xcodebuild -project Fluid.xcodeproj -scheme Fluid -destination 'platform=macOS' build CODE_SIGNING_ALLOWED=NO
   

6. (Optional) Install pre-commit hook to prevent accidental team ID commits:

   cp scripts/check-team-id.sh .git/hooks/pre-commit
   chmod +x .git/hooks/pre-commit
   

Pull Request Guidelines

• Keep changes focused and atomic - one feature or fix per PR
• Create an issue before raising a PR so the work is easy to track and validate before review starts
• Discuss before raising a PR for anything non-trivial so it is easier to merge and review tradeoffs up front
• Explain the pros and cons of your approach when discussing larger changes
• Follow the PR template when opening your pull request
• Update documentation if adding new features
• Test thoroughly on your machine before submitting
• Never commit personal team IDs or API keys to project.pbxproj
• Check git diff before committing to ensure no personal settings leaked in

Connect

Follow development updates on X: @ALTIC_DEV

Run integration dictation test

xcodebuild test -project Fluid.xcodeproj -scheme Fluid -destination 'platform=macOS'

If you run into a test bundle load error related to code signing/Team ID, run without overriding code signing flags (the command above), or explicitly:

xcodebuild test -project Fluid.xcodeproj -scheme Fluid -destination 'platform=macOS' CODE_SIGN_STYLE=Automatic

CI uses unsigned builds:

xcodebuild test -project Fluid.xcodeproj -scheme Fluid -destination 'platform=macOS' CODE_SIGNING_REQUIRED=NO CODE_SIGNING_ALLOWED=NO

Privacy & Analytics

FluidVoice is designed to be local-first, but it includes optional anonymous analytics, solely to keep track of FV usage and future feature development.

What this helps with

• Feature usage patterns (for example dictation, command mode, and write mode)
• Product reliability and performance tuning
• Anonymous release health signals to prioritize fixes and improvements

What is collected

• App version, build, and macOS version
• Low-cardinality feature/config flags (for example app mode and major settings)
• Approximate usage ranges (not exact values)
• High-level success/error outcomes

What is not collected

• Your voice, Raw audio or transcribed text
• Selected text, prompts, or AI-generated responses
• Terminal commands or outputs
• Window titles, file names/paths, clipboard content, or typed content
• or ANYTHING personal or private information.

How to disable

Analytics are enabled by default. You can disable or re-enable them at any time from:

Settings → Share Anonymous Analytics

Why this exists

Anonymous telemetry helps us understand what breaks, where performance can be improved, and which features matter most without collecting personal content. This helps us continue building what users want and if we should even continue developing features for you.

License History

• Versions before 2026-02-23: Apache License 2.0
• Versions on and after 2026-02-23: GNU General Public License v3.0 (GPLv3)

License

From 2026-02-23 onward, this project is licensed under the GNU General Public License, Version 3.0 (GPLv3).

Versions published before this date were licensed under Apache License 2.0.

---