{"version":1,"type":"story","url":"https://digestai.news/story/hugging-face-releases-anylanguagemodel-1-0-for-swift-apps-on-apple-pla","json":"https://digestai.news/story/hugging-face-releases-anylanguagemodel-1-0-for-swift-apps-on-apple-pla.json","markdown":"https://digestai.news/story/hugging-face-releases-anylanguagemodel-1-0-for-swift-apps-on-apple-pla.md","slug":"hugging-face-releases-anylanguagemodel-1-0-for-swift-apps-on-apple-pla","headline":"Hugging Face releases AnyLanguageModel 1.0 for Swift apps on Apple platforms","summary":"Hugging Face announced AnyLanguageModel 1.0, a Swift library that lets iOS, macOS, visionOS and Linux apps use local or hosted language models through a shared Foundation Models API. The update adds tool calling, guided generation (including nested @Generable types) and image support for all providers, and moves API‑key handling to closures for short‑lived tokens.\n\nThe library now works with OpenAI’s Chat Completions, Anthropic’s Messages, Google’s Gemini, Ollama, Private Cloud Compute, Hugging Face Inference Providers and OpenRouter endpoints. Local models can run via llama.cpp, MLX or Core ML, while larger open‑weight models can be accessed through the Inference Providers layer. Popular downloadable models listed on the Hub include Gemma 4 E2B (≈3.4 GB), LFM2.5 2.6B (≈1.7 GB), Qwen3.8 27B (≈16 GB) and GLM‑5.3 (≈1.5 TB). Swift‑huggingface 1.0 and swift‑xet 1.0 also ship to accelerate model downloads. Migration guidance is provided for developers updating existing apps.","keyPoints":["AnyLanguageModel 1.0 adds tool calling, guided generation, and image support across all providers.","Supports OpenAI, Anthropic, Google Gemini, Ollama, Private Cloud Compute, Hugging Face Inference and OpenRouter via unified API.","Popular downloadable models include Gemma 4 E2B (3.4 GB), LFM2.5 2.6B (1.7 GB), Qwen3.8 27B (16 GB) and GLM‑5.3 (1.5 TB)."],"whyItMatters":"The release lets iOS, macOS and visionOS developers run or call a wide range of language models with a single Swift API, simplifying AI integration and expanding on‑device capabilities for millions of apps.","category":{"slug":"agents","name":"Agents & Tools","url":"https://digestai.news/category/agents"},"entities":{"companies":["Hugging Face","Apple","OpenAI","Anthropic","Google","Alibaba"],"models":["Gemma 4 E2B","LFM2.5 2.6B","Qwen3.8 27B","GLM-5.3"],"people":[]},"firstPublishedAt":"2026-10-20T00:00:00Z","updatedAt":"2026-10-20T00:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"Hugging Face","title":"AnyLanguageModel 1.0: One Year of Local Models on Apple Platforms","url":"https://huggingface.co/blog/anylanguagemodel-1-0","publishedAt":"2026-10-20T00:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"Reflection AI's Beam Model Saga","url":"https://digestai.news/thread/reflection-unveils-beam-an-open-model-it-says-rivals-glm-5-2-on-reasoning-while","storyCount":3},"cite":{"text":"Digest AI, \"Hugging Face releases AnyLanguageModel 1.0 for Swift apps on Apple platforms\", 20 October 2026, https://digestai.news/story/hugging-face-releases-anylanguagemodel-1-0-for-swift-apps-on-apple-pla","publisher":"Digest AI","title":"Hugging Face releases AnyLanguageModel 1.0 for Swift apps on Apple platforms","datePublished":"2026-10-20T00:00:00Z","url":"https://digestai.news/story/hugging-face-releases-anylanguagemodel-1-0-for-swift-apps-on-apple-pla"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}