{"slug": "show-hn-we-ve-built-a-standalone-apple-watch-app-running-llms-offline-locally", "title": "Show HN: We've built a standalone Apple Watch app running LLMs offline, locally", "summary": "NobodyWho has released the world's first standalone Apple Watch app that runs AI models locally and offline without requiring an iOS companion app. The app supports small language models like Prism ML's LFM2.5 and Liquid AI's 1-bit Bonsai, while a companion Vision Pro app runs larger thinking models such as Qwen3.", "body_md": "# Apple Watch & Vision Pro apps\n\nWe are excited to announce the release of our first two apps! Both are personal AI assistants that run completely offline, powered by the NobodyWho inference engine.\n\n## Apple Watch\n\nThis is a standalone Apple Watch app, meaning it works without any iOS companion app. It's the **world's first** Apple Watch app that can run an AI model locally, completely offline!\nThe idea was to create an app that could answer simple questions on the go, off the grid, in a single shot, which is why we didn't bother persisting conversations.\n\nSince computing power is limited on the Apple Watch without Metal acceleration, we are restricted in the models we can run. Currently, we support models from [Prism ML](https://prismml.com/) and [Liquid AI](https://www.liquid.ai/), such as the LFM2.5 with 700 million parameters (Apple Watch Ultra) and the 1-bit Bonsai with 1.7 billion parameters (Apple Watch Series 11).\n\nAs smaller models get smarter over time, the door opens to better answers on the Apple Watch, and we are very excited for the new small language models to come!\n\nLinks:\n\n## Vision Pro\n\nFollowing the path set by the Apple Watch app, we decided to build an app that can run large, capable, thinking models. Streamed answers render beautifully thanks to the [LLMStream](https://github.com/synth-inc/LLMStream) library. Models from Prism ML and Liquid AI are supported as well, along with Qwen3, a thinking model.\n\nLinks:\n\n## Who We Are\n\nWe're NobodyWho, a local inference library that enables running small models on edge devices for [Swift](https://docs.nobodywho.ooo/swift/) and several other languages.\nWe value open-source code, control over your models, solid software engineering, standardization, and making simple things simple.\n\nAll of which is missing in today's AI world. Running a model with us is as easy as:\n\n``` python\nimport NobodyWho\n\nlet chat = try await Chat.fromPath(modelPath: \"./model.gguf\")\nlet response = try await chat.ask(\"What is the capital of Denmark?\").completed();\n\n// The capital of Denmark is Copenhagen.\n```\n\nIf you value the same things, come and [become a contributor](https://github.com/nobodywho-ooo/nobodywho) or just [download and test our library](https://docs.nobodywho.ooo/).\n\nPublished Jun 4, 2026", "url": "https://wpnews.pro/news/show-hn-we-ve-built-a-standalone-apple-watch-app-running-llms-offline-locally", "canonical_source": "https://nobodywho.ooo/posts/apple-watch-vision-pro-apps/", "published_at": "2026-06-04 11:35:52+00:00", "updated_at": "2026-06-04 11:48:49.904883+00:00", "lang": "en", "topics": ["large-language-models", "ai-products", "ai-startups", "ai-tools", "artificial-intelligence"], "entities": ["NobodyWho", "Prism ML", "Liquid AI", "Apple Watch", "Vision Pro", "LFM2.5", "Bonsai", "Qwen3"], "alternates": {"html": "https://wpnews.pro/news/show-hn-we-ve-built-a-standalone-apple-watch-app-running-llms-offline-locally", "markdown": "https://wpnews.pro/news/show-hn-we-ve-built-a-standalone-apple-watch-app-running-llms-offline-locally.md", "text": "https://wpnews.pro/news/show-hn-we-ve-built-a-standalone-apple-watch-app-running-llms-offline-locally.txt", "jsonld": "https://wpnews.pro/news/show-hn-we-ve-built-a-standalone-apple-watch-app-running-llms-offline-locally.jsonld"}}