OPEN-SOURCE LOCAL TOOL
From an example
to your own data.
The interactive demo now lives on the homepage. To measure your own endpoint, run Inferpico on your computer.
Open homepage demo01 / Run on your computer
Real measurements.
Your environment.
You need Python 3.10 or later. Install the release into a virtual environment, then use the inferpico command or local browser interface. The installed tool has no third-party runtime dependencies.
Read the GitHub setup guide for macOS, Linux, and Windows.Version 0.2.0 measures full-response latency, reported tokens, and a cost estimate based on your rates. It does not measure TTFT or evaluate answer quality.
Installation commands
macOS / Linux: open a terminal in a new folder. These commands install v0.2.0 from GitHub into a virtual environment.
python3 -m venv .venv
.venv/bin/python -m pip install "https://github.com/newsbruno/inferpico/releases/download/v0.2.0/inferpico_check-0.2.0-py3-none-any.whl"
.venv/bin/inferpico demo
.venv/bin/inferpico serveAfter the last command, open http://127.0.0.1:8765. The offline demo requires no key. Live requests send your prompt and any configured API key to the endpoint you choose and may incur charges. Review the security guide before configuring an endpoint. There is no hard dollar spending cap.
Code and references.
Documentation and local tool in English.