Ollama Now Speaks: Your Computer Can Read Aloud to You
Listen to AI responses instead of reading them—hands-free and eyes-free.
Ollama, a popular free app that lets you run AI models on your own computer (instead of the cloud), just added a text-to-speech feature for macOS. Now, when you ask an AI model something, it can read the answer out loud to you. This is perfect if you want to listen while doing other things, or if you have trouble reading on screens. It's like having a personal assistant read your emails or articles to you. The feature also lets other apps use this speech generation through a standard interface, so developers can easily add voice to their tools.
In addition, Ollama now supports a family of small AI models called d3, which are designed to make decisions. You can give them text, structured data, or even an image, and they can classify it, answer yes/no questions, or score options. These models come in four sizes (2B, 4B, 9B, and 27B parameters), so you can pick one that fits your computer's power. They're useful for things like sorting customer emails or flagging content. Because they're small, they run fast and don't need an internet connection.
The catch? The speech feature is only available on macOS for now, so Windows and Linux users will have to wait. Also, while these tools are powerful, they require some technical know-how to set up. But the direction is clear: AI is becoming more accessible and integrated into our daily computing, with voice and decision-making capabilities that were once only in big data centers now on your laptop.
- Ollama now reads AI responses aloud on Macs, so you can listen instead of read.
- New small AI models can make quick decisions like sorting or scoring, running locally on your computer.
- The speech feature is Mac-only for now, and setup may require some tech skills.
Why It Matters
You can listen to AI answers while multitasking, and small decision models make smart apps faster and more private.