Phi-4 Family — local AI model by Microsoft

Written by Jakub Rusinowski · Last updated January 6, 2025

Microsoft's latest model focusing on high-quality synthetic data training. Phi-4 14B outperforms many models twice its size in reasoning benchmarks.

Hugging Face

Variants

The smallest Phi-4 Family variant needs about 9 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.

Model	VRAM at Q4	VRAM	Context	Run it
Phi-4 (14B) → 14B		~9.3 GB	16,000	`ollama run phi4`

Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.

How to run Phi-4 Family locally

Install Ollama, then pull the tag.

$ollama run phi4

Pick a size above for its own VRAM figure, speed estimate and install command.

Download Phi-4 Family — GGUF Quantizations

Pick a quantization and open it in LM Studio, Ollama, or Jan, or download the raw .gguf file directly. Quant list and sizes resolved from Hugging Face.

Phi-4 (14B) — GGUF quants · bartowski/phi-4-GGUF

Quant	Size	Download (.gguf)
Q3_K_M	5.97 GB (est.)	phi-4-Q3_K_M.gguf
Q4_K_M	8.45 GB (est.)	phi-4-Q4_K_M.gguf
Q5_K_M	9.92 GB (est.)	phi-4-Q5_K_M.gguf
Q6_K	11.48 GB (est.)	phi-4-Q6_K.gguf
Q8_0	14.88 GB (est.)	phi-4-Q8_0.gguf

Download in LM Studio: lms get bartowski/phi-4-GGUF

Want this model on your phone? You can run it on your desktop with LM Studio and chat from your iPhone or iPad over an encrypted link — see Run LM Studio Models on Your Phone (LM Link).

Licence

MITCommercial use permitted

Commercial use permitted. No usage restrictions beyond attribution.

Applies to: Phi-4 (14B)

Recommended GPU

The cheapest catalogued GPU that runs Phi-4 Family locally (min 9 GB VRAM) is the Intel Arc B570 (10 GB).

Affiliate disclosure: Some links on this page are affiliate links — if you buy through them, LLM Configurator may earn a commission at no extra cost to you. As an Amazon Associate, LLM Configurator earns from qualifying purchases.

Intel Arc B570 10GB

10 GB VRAM · 150 W board power

2026 prices are volatile — check the current listing.

Check price on Amazon

Can I run Phi-4 Family on my GPU?

Phi-4 Family — frequently asked questions

How much VRAM does Phi-4 Family need?

Phi-4 Family needs about 9 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Phi-4 (14B) (9 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.

Can I run Phi-4 Family on an RTX 4090 (24 GB)?

Yes — Phi-4 Family runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.

What quantization should I use for Phi-4 Family?

Q4_K_M is the best balance of quality and VRAM for Phi-4 Family in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.

How do I run Phi-4 Family with Ollama?

Install Ollama, then run: ollama run phi4. This downloads Phi-4 Family and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.

Official links

Hugging Face