Runtime comparison - verified
Microsoft Foundry Local vs Ollama
Microsoft Foundry Local (Microsoft) and Ollama, compared field by field from their official docs as of October 18, 2026. Their docs differ on 8 feature points, listed below; the table also shows versions, install methods and links.
Key differences
- License: Ollama is open source (MIT); Microsoft Foundry Local is not (Microsoft Software License Terms for the CLI, MIT for the SDK).
- Interface: Microsoft Foundry Local lists Python library and Ollama does not; Ollama lists Desktop app and Docker and Microsoft Foundry Local does not.
- Backends: Microsoft Foundry Local lists NPU and Other and Ollama does not; Ollama lists ROCm, Metal and Vulkan and Microsoft Foundry Local does not.
- Model formats: Microsoft Foundry Local lists ONNX and Ollama does not; Ollama lists GGUF and Safetensors and Microsoft Foundry Local does not.
- Anthropic-compatible API: Microsoft Foundry Local does not state it in its docs; Ollama documents it.
- Structured outputs: Microsoft Foundry Local does not state it in its docs; Ollama documents it.
- Vision (image input): Microsoft Foundry Local does not state it in its docs; Ollama documents it.
- Multi-GPU: Microsoft Foundry Local does not state it in its docs; Ollama documents it.
Generated from the fields below that differ. Versions, dates and links are left to the table.
Side by side
Every field we track, with the source linked on each value. Rows marked Differs are where the two runtimes' docs differ.
| Field | Microsoft Foundry Local | Ollama |
|---|---|---|
| VendorDiffers | Microsoft | Ollama |
| LicenseDiffers | Microsoft Software License Terms for the CLI, MIT for the SDK | MIT |
| Open sourceDiffers | No | Yes |
| Latest releaseDiffers | v2.1.0 | v0.40.6 |
| Release dateDiffers | September 30, 2026 | October 16, 2026 |
| InterfaceDiffers | CLI, Server, Python library | CLI, Server, Desktop app, Docker |
| Platforms | Windows, macOS, Linux | macOS, Windows, Linux |
| BackendsDiffers | CPU, CUDA, NPU, Other | CPU, CUDA, ROCm, Metal, Vulkan |
| Model formatsDiffers | ONNX | GGUF, Safetensors |
| OpenAI-compatible API | YesOptional OpenAI-compatible local server started with `foundry server start` (or `start_web_service()` in the SDK); no fixed default port is documented, `--port <port>` sets one and `foundry server status` confirms it | Yeshttp://localhost:11434/v1 |
| Anthropic-compatible APIDiffers | Not stated | Yes |
| Tool calling | Yes | Yes |
| Structured outputsDiffers | Not stated | Yes |
| Vision (image input)Differs | Not stated | Yes |
| Multi-GPUDiffers | Not stated | Yes |
| MCP client | Not stated | Not stated |
| Install methods listedDiffers | windows; macos; python; javascript | macOS; Windows; Linux; Docker (CPU only) |
| Run exampleDiffers | foundry run qwen2.5-0.5b | ollama run gemma4 |
| Audience, per the vendorDiffers | Developers who want to ship on-device AI inside their own applications using native SDKs, a curated catalog of optimized ONNX models and automatic NPU, GPU or CPU acceleration for a single user per device. | Developers who want to download and run open models from the command line, with a local REST API on port 11434 and built-in launchers for coding agents such as Claude Code and Codex. |
| HomepageDiffers | learn.microsoft.com/en-us/azure/foundry-local | ollama.com |
| Source repositoryDiffers | github.com/microsoft/Foundry-Local | github.com/ollama/ollama |
Install both
Copied verbatim from the page linked under each command.
Microsoft Foundry Local
windows
winget install Microsoft.FoundryLocalSource: learn.microsoft.com/en-us/azure/foundry-local/reference/reference-cli
macos
brew tap microsoft/foundrylocal
brew install foundrylocalSource: learn.microsoft.com/en-us/azure/foundry-local/reference/reference-cli
javascript
npm install foundry-local-sdkOllama
Docker (CPU only)
docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollamaSource: docs.ollama.com/docker
FAQ
What is the difference between Microsoft Foundry Local and Ollama?
Going by their official docs: License: Ollama is open source (MIT); Microsoft Foundry Local is not (Microsoft Software License Terms for the CLI, MIT for the SDK). Interface: Microsoft Foundry Local lists Python library and Ollama does not; Ollama lists Desktop app and Docker and Microsoft Foundry Local does not. Backends: Microsoft Foundry Local lists NPU and Other and Ollama does not; Ollama lists ROCm, Metal and Vulkan and Microsoft Foundry Local does not. Model formats: Microsoft Foundry Local lists ONNX and Ollama does not; Ollama lists GGUF and Safetensors and Microsoft Foundry Local does not. The table lists 4 more.
Do Microsoft Foundry Local and Ollama both have an OpenAI-compatible API?
Microsoft Foundry Local: yes. Optional OpenAI-compatible local server started with `foundry server start` (or `start_web_service()` in the SDK); no fixed default port is documented, `--port <port>` sets one and `foundry server status` confirms it. Ollama: yes. http://localhost:11434/v1.
Which platforms do Microsoft Foundry Local and Ollama run on?
Microsoft Foundry Local: Windows, macOS, Linux. Ollama: macOS, Windows, Linux. Both list Windows, macOS and Linux.
Are Microsoft Foundry Local and Ollama open source?
Microsoft Foundry Local is not open source (Microsoft Software License Terms for the CLI, MIT for the SDK). Ollama is open source (MIT).
Do Microsoft Foundry Local and Ollama support tool calling and MCP?
Tool calling: Microsoft Foundry Local yes, Ollama yes. MCP client: Microsoft Foundry Local not stated, Ollama not stated. "Not stated" means the docs do not say.
What are the latest versions of Microsoft Foundry Local and Ollama?
As of October 18, 2026: Microsoft Foundry Local v2.1.0 (September 30, 2026) and Ollama v0.40.6 (October 16, 2026).
More comparisons
Newsletter
Local AI, every week
One email per week. New model releases, runtime updates and what they mean for what you build.
Keep exploring
Open-weight models you can run
Download sizes, licenses and the run command for each model.
All local LLM runtimes
Every runtime in one feature matrix, with sources.
Picking a runtime for coding agents
Ollama, LM Studio, vLLM and llama.cpp for agent workloads.
Setup guides
Point your coding tools at the model you picked.