Foundry Local
Microsoft's on-device model runtime with an OpenAI-compatible endpoint.
Overview
Foundry Local runs optimized open models on Windows and macOS, picking the best variant for your CPU, GPU or NPU. It exposes an OpenAI-compatible API and a CLI for downloading and serving models.
- Hardware-aware model selection
- OpenAI-compatible API
- CLI and SDKs
- Model cache
Our verdict
What we like
- Simple setup on Windows
- Uses NPUs where available
Watch out for
- Smaller model catalog than Ollama
Works with these providers
Keys from these providers work in this app. Usage is billed by the provider at its list price.
BYOK Key Safety Score
Key handling in Foundry Local
Grade A
Your key is stored on your device and requests go directly to the provider.
- Key storage40/40
Key stays on your device; requests go direct to the provider.
- Auditable code25/25
Open source: anyone can verify how keys are handled.
- Self-hosting8/15
Runs as a local app, but not as a self-hosted service.
- Offline option10/10
Supports local models, so no key is needed at all for private work.
- Provider choice2/10
1 provider option, which lets you pick the provider whose data terms you trust.
Whatever the score, give each app its own key, set a monthly spend limit with your provider, and revoke keys you no longer use. Read the key-safety checklist.
Foundry Local FAQ
- Does Foundry Local support bring your own key?
- Yes. Foundry Local accepts your own API key from OpenAI-compatible. Usage is billed by the provider at its API rates.
- Where does Foundry Local store my API key?
- On your device. Requests go directly from Foundry Local to the provider. Its BYOK Tools Key Safety Score is 85/100.
- Is Foundry Local free?
- Free: Free. API usage is billed separately by your model provider.
- Is Foundry Local open source?
- Yes. The source code is at https://github.com/microsoft/Foundry-Local.
- Can Foundry Local run local models?
- Yes. It can connect to local runtimes such as Ollama or LM Studio, so no API key is needed for private or offline use.
- Which platforms does Foundry Local support?
- Windows, macOS, CLI.
Community reviews
No reviews yet.
Alternatives to Foundry Local
All local ai →llama.cpp
Local AI
C and C++ LLM inference engine with a built-in server and web UI.
Ollama
Local AI
The simplest way to run open LLMs locally with one command.
llamafile
Local AI
Run an LLM from a single executable file on any OS.
Jan
Local AI
Open-source ChatGPT alternative that runs 100% offline.
LocalAI
Local AI
Self-hosted, drop-in OpenAI API replacement for local models.
LM Studio
Local AI
Desktop app to discover, download and run local LLMs.