Edgeinfer-eval is a real-time CPU-based object detection pipeline built in Nim using ONNX Runtime and OpenVINO, achieving 130–170 fps inference on an Intel Core i5-11400 with standard USB cameras. The evaluation build is free for testing, education, and research (not commercial use), with closed source and compiled binaries; the developer requests CPU benchmark feedback and global-shutter camera results to assess image quality improvements.