Hugging Face Releases AnyLanguageModel 1.0, Unifying Local and Hosted AI Models on Apple Platforms

Hugging Face has launched AnyLanguageModel 1.0, a drop‑in replacement for Apple’s Foundation Models framework that lets developers use any model—local via llama.cpp, MLX, Core ML, or hosted services—through a single API. The update adds tool‑calling, guided generation, image prompt support, and compatibility with OpenAI, Anthropic, Google Gemini, Ollama, and Hugging Face inference endpoints. It also introduces swift‑huggingface 1.0 and swift‑xet 1.0 for accelerated model downloads in Swift apps.