The rapid expansion of the Internet of Things (IoT) has significantly raised concerns over secure identification and authentication of IoT devices. While physical unclonable function (PUF)-based anticounterfeiting cryptography shows promise, implementing multi-factor authentication (MFA) system with resource-constrained PUF device remains a significant challenge. Here, we demonstrate a multidimensional-encoded optical PUF fabricated with multi-color quantum dots (mQDs), resilient to machine learning attacks, for use in anticounterfeiting MFA. Randomly distributed mQDs in a periodic nanostructure fabricated via nanoimprint lithography generate spatially chaotic, unpredictable multiple security keys when ultraviolet (UV) light is only illuminated. Photoluminescence measurements revealed irregular mQD emission, where disordered distribution-induced Förster resonant energy transfer produces unpredictable color patterns. Our PUF-induced multiple keys were validated through advanced PUF metrics like uniformity, uniqueness, correlation factor, entropy, and even resilience to machine learning attack. We further demonstrate efficient implementation of a cryptography protocol with MFA system for IoT applications using mQDs-based optical PUFs. Material systems with physical unclonable functions (PUF) are needed for secure authentication. Here, a multidimensional-encoded optical PUF fabricated with multi-color quantum dots is demonstrated, for multi-factor authentication.