{"slug": "running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned", "title": "Running a chat model and an SDXL model together on an iPhone: what I learned building Anvil", "summary": "Anvil, an iPhone app built by a local-model developer, runs Gemma 4 E2B chat via LiteRT LM and an SDXL image model converted to Core ML on the Neural Engine at 1024x1024, with both models never loaded simultaneously so the app fits in memory. The app requires an iPhone 15 Pro or newer, needs no account, and keeps web search and iCloud backup as the only optional network features; version 4.1 is available on public TestFlight with free Anvil Pro for testers.", "body_md": "Long time local model fan here. I’ve been building Anvil, an iPhone app that runs everything on the phone, and wanted to share how it fits.\n\nChat is Gemma 4 E2B through LiteRT LM. Photo and document questions use the same model, so it can read and translate a sign or a menu with no signal.\n\nImages come from an SDXL model converted to Core ML, running on the Neural Engine at 1024x1024, as many as you want, even in airplane mode. The grid above was all made on the phone.\n\nThe memory trick is boring but it works: the two models are never loaded at the same time. That’s also why it needs an iPhone 15 Pro or newer.\n\nNo account, and nobody else sees what you ask. Web search and iCloud backup are optional and are the only parts that touch the network.\n\nVersion 4.1 is on public TestFlight and testers get Anvil Pro free: [Join the Anvil – Offline AI Assistant beta - TestFlight - Apple](https://testflight.apple.com/join/s8UAwXBm)\n\nIf you’ve pushed other diffusion models onto iOS, I’d love to hear what compression settings kept quality up.", "url": "https://wpnews.pro/news/running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned", "canonical_source": "https://forum.level1techs.com/t/running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned-building-anvil/258250#post_1", "published_at": "2026-10-09 22:02:33+00:00", "updated_at": "2026-10-09 22:23:39.846565+00:00", "lang": "en", "topics": ["ai-products", "large-language-models", "generative-ai", "ai-tools"], "entities": ["Anvil", "Gemma 4 E2B", "LiteRT LM", "SDXL", "Core ML", "Apple", "iPhone 15 Pro", "TestFlight"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned", "markdown": "https://wpnews.pro/news/running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned.md", "text": "https://wpnews.pro/news/running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned.txt", "jsonld": "https://wpnews.pro/news/running-a-chat-model-and-an-sdxl-model-together-on-an-iphone-what-i-learned.jsonld"}}