# Apple Foundation Models 45 Hacker News commenters recommended Apple Foundation Models and 35 criticized it, November 2025 to November 2026. It was recommended in 11 of 13 months; owners who said how long report a median of 1 years (4 owners). **Verdict:** Developers praise a free, built-in on-device model that apps can call without extra downloads or memory. Critics say it trails cloud models and struggles with non-English prompts. ## Why people recommend it - Surprisingly capable for its size (6 comments): "I've been somewhat surprised at the quality and reliability of Apple's built-in LLM and have only been limited by the logic I've built around it." ([source](https://news.ycombinator.com/item?id=46209975)) "I've been impressed even with how Siri's been doing with the Apple Foundation Models in MacOS/iOS 27 given how small they are." ([source](https://news.ycombinator.com/item?id=49831548)) "Their FoundationModels are actually pretty good at certain tasks" ([source](https://news.ycombinator.com/item?id=46187739)) - Easy API for developers in apps (5 comments): "The idea of a system wide (and platform wide) on device model being a core part of OS APIs is very appealing." ([source](https://news.ycombinator.com/item?id=48460869)) "Apple exposed the API in a fairly complete way. I integrated their local model in less than a day" ([source](https://news.ycombinator.com/item?id=47851405)) "The ability to tap into on-device foundation models within apps is pretty baller" ([source](https://news.ycombinator.com/item?id=48456767)) - Ships with OS, no download (4 comments): "well-balanced and relatively small local-inference models that ship with every mac" ([source](https://news.ycombinator.com/item?id=49958389)) "If you're on macOS you can try the built in LLM which I think is similar in size." ([source](https://news.ycombinator.com/item?id=48366491)) "without having you download a second LLM and double your memory usage" ([source](https://news.ycombinator.com/item?id=48541973)) - Useful for classification and filtering (3 comments): "The Foundation models keep a surprisingly decent set of data available for fast filtering by a local LLM." ([source](https://news.ycombinator.com/item?id=50120228)) "It's perfectly capable of looking at images and classifying them at the level you'll need for this app." ([source](https://news.ycombinator.com/item?id=48346415)) "that's actually pretty powerful for a use case like this" ([source](https://news.ycombinator.com/item?id=48342590)) ## What people warn about - Weaker than cloud and open models (6 comments): "AFM is still incredibly premature and I'm experimenting with newer models that perform much better." ([source](https://news.ycombinator.com/item?id=47587670)) "for small models they are good" ([source](https://news.ycombinator.com/item?id=48453555)) "Not Apple Foundation Models — unfortunately they're not capable enough (yet) for understanding content" ([source](https://news.ycombinator.com/item?id=46331330)) - Apple-only; splits cross-platform apps (3 comments): "It forces you to essentially maintain two app experiences." ([source](https://news.ycombinator.com/item?id=47589820)) "It's only on Apple devices this API is used, so they split up the market by not letting developers use the same system if they want things to work on iOS" ([source](https://news.ycombinator.com/item?id=48538080)) "There is no major opportunity for developers on Apple's platforms when they can just rug pull you as they please." ([source](https://news.ycombinator.com/item?id=46210650)) - Small context window, weak chat (3 comments): "they’re not made for conversation, so keep your expectations down in chat mode." ([source](https://news.ycombinator.com/item?id=47625212)) "seem great on paper until you see the 4k context window" ([source](https://news.ycombinator.com/item?id=47921597)) "It's not very good at world-knowledge recall, but can describe the submitted image well enough." ([source](https://news.ycombinator.com/item?id=48462451)) ## Alternatives hackers recommend - [Claude Haiku](https://hackerapproved.com/p/claude-haiku.md): 218 recommend - [OpenAI Luna](https://hackerapproved.com/p/luna.md): 218 recommend - [Ministral](https://hackerapproved.com/p/ministral.md): 18 recommend - [Microsoft Phi](https://hackerapproved.com/p/phi.md): 16 recommend - [LFM2.5 2.6B](https://hackerapproved.com/p/lfm2-5-2-6b.md): 14 recommend ## Cite as Hacker Approved, "Apple Foundation Models", https://hackerapproved.com/p/apple-foundation-models/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).