Kuzco is a Swift package for integrating large language models (LLMs) directly into iOS, macOS, and Mac Catalyst apps. Built on `llama.cpp`, it offers customizable prompts, flexible tuning, and async/await-friendly APIs for on-device AI.
Hey everyone!
I built Kuzco for two big reasons:
1. Developers can cut costs by skipping cloud APIs and running AI locally on‑device, completely free.
2. I was done sending my data (and users’ data) to third‑party servers.
Kuzco is a straightforward Swift wrapper for llama.cpp, letting you plug local models into iOS and macOS apps with just a few lines of code. It’s fully open source, so dive in, tweak away, and if you use it (or even plan to) please drop me a link on X (Twitter), so I can cheer you on!
Very cool! Love that Kuzco makes it easier to bring LLMs directly into Apple platforms. The async/await-friendly APIs and on-device support sound especially useful—looking forward to trying it out!
Kuzco is a useful open-source Swift package that enables running LLMs locally on iOS and macOS, offering customizable features and user-friendly APIs for seamless on-device AI integration.
Async/await support is chef’s kiss, and the llama.cpp backbone
Thank you for your contribution to the open source community
best thing ever for web-developers!!! Finally multi-modal came to APP world
Kuzco makes on-device LLM integration feel native. Swift-first, async-friendly, and built on llama.cpp finally, customizable AI without the latency tax.
Finally. This was the product I had been waiting for for a long time, and which the global community had been waiting for. Congratulations.
Running AI models locally with just a Swift wrapper? That’s genius—no more sending user data everywhere, ngl this is a game-changer for privacy-first apps. Makers nailed it!
Congrats to your launch! Which offline model do you think is the best for your scenario?
