QuarkAI Unveils TinyML Model Store and Wins $40M Series B for Edge AI

Tech · 6 min read

QuarkAI closed a $40 million Series B to grow its TinyML marketplace and developer tooling that enables companies to deploy compact AI models on constrained devices. The startup launched QuarkHub, a model store featuring quantized vision and audio models optimized for microcontrollers and mobile SoCs.

QuarkHub includes certification badges for latency, memory footprint, and on-device privacy compliance. The company provides a toolchain that automates post-training quantization and adapts models to hardware-specific NPUs. Customers can test models on a virtual edge runtime in the cloud before deployment.

The funding will be used to broaden hardware partnerships, add enterprise features for fleet management of models, and expand regional offices to support automotive and industrial IoT customers. QuarkAI says its offering is aimed at companies prioritizing offline inference and data minimization.