min-llm-server-client 0.3.10

A minimal API server for local HuggingFace LLMs

Minimal LLM Server, for API calls

The simplest possible Python code for running local LLM inference as a REST API server (with a simple client).

This package lets you start an inference server for Hugging Face–compatible models (like LLaMA, Qwen, GPT-OSS, etc.) on your own computer or server, and make it accessible to applications via HTTP.

See the Tutorial page for extented info.

Installation by pip

From PyPI (recommended):

pip install min-llm-server-client

From source:

git clone https://github.com/afshinsadeghi/min_llm_server_client.git
cd min_llm_server_client
pip install .

---

Usage

Starting the Server

After installation, you can launch the server with the provided CLI entrypoint:

min-llm-server --model_name meta-llama/Llama-3.3-70B-Instruct --max_new_tokens 100 --device cuda:0

Options:

• --model_name : Hugging Face model name or local path suggested models: openai/gpt-oss-20b openai/gpt-oss-120b meta-llama/Llama-3.3-70B-Instruct meta-llama/Llama-3.1-8B Qwen/Qwen3-0.6B Qwen/Qwen2-VL-72B-Instruct-AWQ deepseek-ai/DeepSeek-R1-Distill-Qwen-32B

  or it can be local model with /path/to/model

• --max_new_tokens : maximum number of tokens to generate in response.

• --device : cpu, cuda:0, cuda:1, etc. if device is not given it finds available GPU core and uses them and if not available it uses CPU instead.

Example (CPU run):

min-llm-server --model_name openai/gpt-oss-20b --max_new_tokens 50 --device cpu

---

Sending Queries

Once the server is running (default: http://127.0.0.1:5000/llm/q), you can query it with curl or Python.

Curl:

curl -X POST http://127.0.0.1:5000/llm/q \
  -H "Content-Type: application/json" \
  -d '{"query": "What is Earth?", "key": "key1"}'

Python client:

from min_llm_server_client.local_llm_inference_api_client import send_query

response = send_query("What is the capital of France?", user="user1", key="key1")
print(response)

---

Performance notes

• Running LLaMA 3.1 8B:
  • Intel CPU → ~30 seconds per request, ~2.4 GB RAM
  • A100 GPU → <1 second per request, ~34 GB GPU memory, ~4.8 GB CPU RAM

---

Project Structure

min_llm_server_client/
├── src/
│   ├── local_llm_inference_api_client.py
│   ├── local_llm_inference_server_api.py
│   └── ...
└── README.md

---

License

This project is open source under the Apache 2.0 License.

---

Author

Afshin Sadeghi
📧 [e-mail]
🔗 GitHub
🔗 Google Scholar
🔗 LinkedIn