Haplo AI
A local AI assistant for iOS and macOS that runs Llama, Qwen and other models on-device, with offline chat and no account required.
Haplo AI is a native AI assistant for iOS and macOS that keeps conversations on your device. It's for people who want to ask questions, work with code or examine photos without sending that material to a hosted assistant. After downloading a model, you can use it offline without an account. Prompts and replies stay in the app.
Its model selection includes Llama, Qwen, Phi, Gemma, Mistral and SmolLM. You can keep several models and choose a different one for each conversation, using a smaller model for quick answers or a larger one for more demanding work. Replies appear as the model generates them, so you don't have to wait for a complete answer before reading. The app runs natively on Apple Silicon and offers models that fit the device's RAM; the Mac app can load larger models than an iPhone can handle.
Chat isn't its only function. A vision model can examine a photo and answer questions about it without uploading the image. Stable Diffusion generates images locally, with prompts and results kept on the device, and the iPhone app also supports on-device video generation. For coding, it can write and debug code without a connection, so you can work offline without submitting a codebase to a remote service.
Model downloads need a connection to the model host. Local chat and generation then run on your own hardware. Optional online tools, such as web search, need network access.
