How to use from
llama.cpp
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
Use Docker
docker model run hf.co/WorkOffline/SmolLM3-3B-GGUF:Q4_K_M
Quick Links

SmolLM3-3B-GGUF

An unchanged copy of SmolLM3 3B GGUF files, kept here so the WorkOffline app can download them from one place, pinned to an exact revision and checked against SHA-256.

File here Copied from Upstream revision Size (bytes) SHA-256
SmolLM3-3B-Q4_K_M.gguf ggml-org/SmolLM3-3B-GGUF/SmolLM3-Q4_K_M.gguf 4965cb60b150737b68a0408c36aeefb65078f894 1915305312 8334b850b7bd46238c16b0c550df2138f0889bf433809008cc17a8b05761863e

The bytes are identical to the upstream files; only the file name may differ. All credit for the model goes to its authors (HuggingFaceTB) and for the GGUF conversion to ggml-org. Use of the files is governed by the original licence.

Downloads last month
71
GGUF
Model size
3B params
Architecture
smollm3
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for WorkOffline/SmolLM3-3B-GGUF

Quantized
(139)
this model