Speech Swift bundles many open speech models for Apple silicon: speech recognition (Qwen3-ASR, Whisper, Parakeet, Nemotron streaming and others), text-to-speech (Qwen3-TTS, Kokoro, CosyVoice and others), forced alignment, voice activity detection, diarization and speech-to-speech, running on MLX Swift and Core ML.
Use it from the speech command line tool, a local HTTP and WebSocket server with OpenAI-compatible transcription and realtime endpoints, or as Swift packages where each model is its own target. Some bundled models carry non-commercial licences; check each model's page.
It needs macOS 15 or iOS 18 on Apple silicon. It's free and open source (Apache 2.0), and audio stays on your device.
How to Get It
brew install speechOfficial downloads only. MLX never re-hosts an app, and never runs a command for you.
MLX Score
Ranked against the voice & audio listings, from real public data only. How the score works.
Pairs Well With
WhisperKitVoice & audioInformation
- Developer
- soniqo
- Category
- Voice & audio
- Price
- Free
- Your data
- Runs on your Mac
- Licence
- Apache-2.0
- Mac
- macOS 15.0 or later
- iPhone and iPad
- iOS 18.0 or later
- Needs
- M1 or later
- Links
- Website · Docs · GitHub · Hugging Face
- Facts checked
- 2026-11-03
Sources
You Might Also Like
SunoVoice & audio
whisper.cppVoice & audio
Wispr FlowVoice & audio
ElevenLabsVoice & audio
HandyVoice & audio
Otter.aiVoice & audio
AikoVoice & audio
MonologueVoice & audio
SuperwhisperVoice & audio
TypelessVoice & audio
KrispVoice & audioSpeech Swift is a product of soniqo; MLX isn’t affiliated with it. Something wrong? Write to support@mlx.app.