Science & Tech · AI · 4 hrs ago
Ollama adds local text-to-speech models to its Mac app
Ollama has added text-to-speech to its Mac app, using Phonon, a family of models developed by Gradium.
Ollama lets people run compatible AI models on their own computers, and on Macs it uses Apple’s MLX framework.
Speech is generated as the AI produces its response, so users can listen without waiting for the full text to be finished.
Phonon supports English, French, German, Spanish and Portuguese, with different voices available depending on the model.
The models are available on Apple Silicon Macs running Ollama 0.40.11 or later, and can be downloaded through Ollama.
Windows and Linux support is expected soon, but no timing is given.
Running models locally can let users use AI without an internet connection or sending data to external servers, though performance depends on the computer’s hardware.
Ollama’s Mac app now supports text-to-speech generation using the Phonon model family.
The feature is available on Apple Silicon Macs running Ollama version 0.40.11 or later.
Phonon models support English, French, German, Spanish and Portuguese, with separate models for each language and a multilingual model.
Ollama says speech is generated in real time, so users can listen while a response is being produced.
Windows and Linux support is planned but was not yet available.
- Who
- Ollama added the feature, using Phonon models developed by Gradium.
- What
- The Mac app now supports generating speech from text.
- When
- Announced on October 26, 2026; available for Ollama 0.40.11 or later.
- Where
- On Apple Silicon Macs.
- Why
- Not stated.
This story does not have two clearly opposing sides.
This story does not have a timeline yet.
- Model family
- Phonon, developed by Gradium
- Supported languages
- English, French, German, Spanish and Portuguese
- Mac requirement
- Apple Silicon and Ollama 0.40.11 or later
- Performance test
- 35% faster speech generation than Gradium’s implementation on a MacBook Pro with M5 Max
What people are saying
Ollama can now turn text into speech.