znyx-inference is an open-source HTTP service designed to serve machine learning models for model-backed guardrail detection. It runs as a sidecar, boots dependency-free, and can serve real ONNX weights on CPU. It is intended for ML engineers who need lightweight, secure inference in their applications.
In the AI space, znyx-inference takes a focused approach. It focuses on serving machine learning models for guardrail detection in applications without complex dependencies. It is built as an open-source project for machine learning engineers. The project is open source (Apache-2.0). znyx-inference is available on the command line and API, and it can be self-hosted.
Behind znyx-inference is zitrino-oss, and it first shipped in 2026. The project is developed in the open on GitHub with 21 commits in the last 90 days. Among its 6 catalogued features are HTTP API, ONNX model support, and guardrail detection.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do