Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Microsoft - Dense - verified

Run Phi-4-mini-reasoning locally

Phi-4-mini-reasoning is a 3.8B dense model from Microsoft with a 131K-token context window, released under MIT (commercial use is allowed); the smallest listed download is 2.2 GB, the 4bit MLX build from mlx-community/Phi-4-mini-reasoning-4bit.

Official weights: microsoft/Phi-4-mini-reasoning

Total params
3.8B
Architecture
Dense
Context
131K
Smallest download
2.2 GB

Fits in

Smallest download at most 85% of memory

Downloads

The official weights and every quantization listed for this model, with file sizes from Hugging Face.

Phi-4-mini-reasoning downloads and file sizes
DownloadFormatSizeRepo
Official weightsBF167.7 GBmicrosoft/Phi-4-mini-reasoning
Q4_K_MGGUF2.5 GBunsloth/Phi-4-mini-reasoning-GGUF
Q8_0GGUF4.1 GBunsloth/Phi-4-mini-reasoning-GGUF
4bitSmallest listedMLX2.2 GBmlx-community/Phi-4-mini-reasoning-4bit

GB is decimal: 1 GB = 10^9 bytes. Add memory for the context and the operating system on top of the file size.

Run it

Copied verbatim from the page linked under each command. Server commands from model cards often assume a multi-GPU node; check the flags before you run them.

Ollama

Terminal
ollama run phi4-mini-reasoning

Source: ollama.com/library/phi4-mini-reasoning

Not sure which runtime to use? Compare local LLM runtimes by platform, API and features.

License

License
MITmit
Commercial use
Commercial use: YesCommercial use is allowed.
Gated download
No. The files download directly.

Worth knowing

Math-focused reasoning tune of Phi-4-mini whose card gives only an April 2025 release month, and Ollama's q8_0 and fp16 tags list a 4K context.

Released (Hugging Face repo creation date).

FAQ

How big is the Phi-4-mini-reasoning download?

The official BF16 weights in microsoft/Phi-4-mini-reasoning are 7.7 GB. The smallest listed download is 2.2 GB, the 4bit MLX build from mlx-community/Phi-4-mini-reasoning-4bit. Plan for more memory than the file size: the context (KV cache) and the operating system need room too.

Can I use Phi-4-mini-reasoning commercially?

Phi-4-mini-reasoning is released under MIT, and commercial use is allowed. The license text is at https://huggingface.co/microsoft/Phi-4-mini-reasoning/blob/main/LICENSE.

How much memory do I need for Phi-4-mini-reasoning?

By this directory's rule of thumb it fits in 8 GB and up: the smallest listed download, 2.2 GB, is at most 85% of 8 GB, which leaves room for the context and the operating system.

How do I run Phi-4-mini-reasoning with Ollama?

Run "ollama run phi4-mini-reasoning", as shown at https://ollama.com/library/phi4-mini-reasoning. The Ollama library lists the phi4-mini-reasoning:3.8b tag at 3.2GB.

What is the context length of Phi-4-mini-reasoning?

131,072 tokens. A longer context needs more memory for the KV cache, so you can run with a shorter one when memory is tight.

Sources

Last verified from the pages below.

Newsletter

New open weights, every week

One email per week. New model releases, open-weight drops and what they mean for what you build.

Keep exploring

Best local AI models

Our curated picks by hardware tier, with benchmarks and the reasoning.

Model API pricing and specs

First-party prices per 1M tokens when you would rather call an API.

AI benchmark leaderboards

Sourced scores, for when you need to compare quality.

Setup guides

Point your coding tools at the model you picked.