How to use from
llama.cppInstall from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf 1bit-MONSTER/phi-4-GGUF# Run inference directly in the terminal:
llama cli -hf 1bit-MONSTER/phi-4-GGUFUse pre-built binary
# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf 1bit-MONSTER/phi-4-GGUF# Run inference directly in the terminal:
./llama-cli -hf 1bit-MONSTER/phi-4-GGUFBuild from source code
git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf 1bit-MONSTER/phi-4-GGUF# Run inference directly in the terminal:
./build/bin/llama-cli -hf 1bit-MONSTER/phi-4-GGUFUse Docker
docker model run hf.co/1bit-MONSTER/phi-4-GGUFQuick Links
phi-4 — GGUF (Q4_K_M)
Microsoft's own Q4_K_M GGUF, re-hosted with measured performance for the 1bit engine on Strix Halo.
Contents
phi-4-Q4_K.gguf(Q4_K_M)
Measured performance (Strix Halo, Vulkan)
pp512: 543 tok/s · tg128: 19.4 tok/s
Running it
1bit serve -m phi-4-Q4_K.gguf --device vulkan
Attribution
- Model and quantization: microsoft/phi-4-gguf, MIT.
- License: MIT, inherited from the base model.
- Downloads last month
- 78
Hardware compatibility
Log In to add your hardware
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for 1bit-MONSTER/phi-4-GGUF
Base model
microsoft/phi-4
Install (macOS, Linux)
# Start a local OpenAI-compatible server with a web UI: llama serve -hf 1bit-MONSTER/phi-4-GGUF# Run inference directly in the terminal: llama cli -hf 1bit-MONSTER/phi-4-GGUF