DeepProve uses Lagrange’s Proof Network (LPN) to assign special-purpose prover sets,
creating a dedicated, modular layer designed for zkML workloads. This setup allows provers to be fine-tuned for different AI models and hardware types — including GPUs and ASICs — enabling faster, more efficient proof generation. By aligning the network with the specific needs of verifiable AI, LPN ensures DeepProve remains scalable, high-performance, and ready for the future as both AI and zero-knowledge technologies advance.
