Recently, Ollama, the open-source platform for running large models locally, announced that its Mac app has officially received a major update, adding a text-to-speech feature for the first time. This upgrade marks a comprehensive step forward for local AI assistants from the traditional "able to read and write" stage to an interactive phase of "being able to speak," and Windows and Linux versions will follow.
In terms of underlying technology and privacy protection, the first supported speech model is the Phonon series developed by Gradium. The model is compact and features advanced technical specifications; users do not need to rely on any cloud services, and can download it directly through Ollama and run it on Mac devices with Apple Silicon chips, making it well suited to real-world scenarios that prioritize data privacy and offline use. Built on macOS, Ollama uses Apple's open-source machine learning framework MLX, fully leveraging the hardware advantages of Apple Silicon's unified memory architecture to greatly improve running efficiency.
Official test data shows that when running Phonon on a MacBook Pro equipped with an M5Max chip, its speech generation speed is 35% faster than Gradium's own implementation, while maintaining an extremely low word error rate comparable to the Hugging Face Transformers reference implementation. In addition, the feature supports real-time speech generation, so users do not need to wait for the AI to output the entire text before hearing the corresponding speech synchronously as the answer is being generated, delivering a more natural and fluent human-like conversation experience.
In terms of specific voices and language support, Phonon currently covers dedicated models for five languages—English, French, German, Spanish, and Portuguese—and also offers a multilingual universal model supporting all five languages, while allowing users to freely choose different accents such as American or British English according to their preference. As a popular open-source project on GitHub, Ollama currently has nearly 185,000 Stars and about 18,600 Forks, continuing to provide developers and ordinary users with a convenient and secure solution for running large models locally.