AnyLanguageModel
Hugging Face's drop-in replacement for Apple's Foundation Models API that also works with other local and cloud model providers.
View on GitHub ↗- Stars
- ★ 952
- Last commit
- 19 Oct 2026
- Maintenance
- Active
- Licence
- Apache-2.0
- Category
- On-device AI SDKs
- GitHub
- huggingface/AnyLanguageModel
Overview
AnyLanguageModel from Hugging Face is a Swift package that works as a drop-in replacement for Apple's Foundation Models framework while supporting many other model providers. You keep the same session API and can switch between Apple's models, local models through Core ML, MLX or llama.cpp, and cloud APIs. It's for app developers who want a choice of which model runs behind their code.
Features
- Supports Apple Foundation Models and Private Cloud Compute
- Local models through Core ML, MLX and llama.cpp (GGUF)
- Ollama, Anthropic, Google Gemini and OpenAI APIs
- Guided generation for on-device models, image inputs and tool calling
- Token usage reporting per response and per session
- Swift package traits keep heavy dependencies optional
Requirements
- Swift 6.3 or later (Xcode 26.4 or later)
- iOS 17.0+, macOS 14.0+, visionOS 1.0+ or Linux
- API keys for any cloud providers you use
Install
Add the package to Package.swift
.package(url: "https://github.com/huggingface/AnyLanguageModel", from: "1.0.0")
Example
Create a LanguageModelSession with an MLX model, ask it to "Explain how rainbows form." and read the token counts from response.usage.
Works with
Listed in
- github.com/huggingface/AnyLanguageModel
- raw.githubusercontent.com/huggingface/AnyLanguageModel/HEAD/README.md
Read the code before you install anything. SwiftAgents doesn't review or endorse the tools listed here. Active: a commit in the last 90 days. Slowing: last commit within the past year. Stale: older than a year, or archived.