Brevis is releasing the ProverNet Whitepaper, detailing the first decentralized zero-knowledge proof generation marketplace.

Source: Brevis

Brevis has today released the ProverNet Whitepaper, introducing its decentralized zero-knowledge proof generation marketplace architecture, which is derived from real-world testing in a mainstream multi-chain protocol production environment. This system addresses a core limitation in existing proof infrastructure—an inability to effectively serve heterogeneous computing workloads requiring different hardware, proof systems, and optimization methods.

ProverNet views proof generation as a two-sided market, where applications submit specific requests and specialized provers bid to fulfill these needs. The market operates through a Truthful Online Double Auction mechanism, designed to handle multiple proof types simultaneously while ensuring honest bidding and optimal resource allocation. ProverNet plans to launch by the end of 2025.

Production Scale Reveals Infrastructure Limitations

Brevis's current infrastructure has generated over 124 million proofs for 98,000 users, spanning protocols such as PancakeSwap, Uniswap, Euler, Linea, and MetaMask, facilitating $224 million in trustless rewards distribution. These deployments have prominently revealed the diversity of proofing work, making it challenging for a single provider architecture to effectively serve.

PancakeSwap's VIP fee discount system requires generating sub-second proofs for individual traders before each trade to verify eligibility. Euler's incentive distribution processes 100,000 addresses every four hours, prioritizing throughput over latency. Linea's ecosystem activity has generated 12.1 million proofs for multi-protocol rewards computation involving 61,902 addresses. Each workload requires different hardware configurations, proof systems (SNARKs vs. STARKs), and performance characteristics.

Michael, CEO and Co-Founder of Brevis, stated, "Existing proof systems are optimized for specific use cases, such as certain Rollups, single virtual machines, or homogeneous workload types. The inception of ProverNet comes from our observation that applications fundamentally require different proofing approaches. A market where specialized provers compete is more efficient than any single operator attempting to serve all use cases."

Market Architecture and TODA Mechanism

ProverNet's architecture treats different proof types as different commodities in an auction. Applications specify proof requirements, including processing types (zkVM execution, data proofs, recursive aggregation), deadlines, maximum fees, and quality parameters. The TODA mechanism calculates the optimal allocation each round, matching heterogeneous requests with suitable proofing capabilities.

This mechanism addresses a unique challenge of the proof generation market. Traditional auction mechanisms are based on the assumption of homogeneous goods, while TODA is able to handle multiple proof types simultaneously. Complex proof tasks are decomposed into subtasks, which are collaboratively executed by different provers.

For example, a zkVM proof may involve chunking on one prover, compressing on another, aggregating on a third, and finally validating packaging on a fourth specialized system.

TODA guarantees several economic properties:

Truthfulness (optimal strategy is to bid honestly)

Budget balance (fees collected exceed fees paid)

Individual rationality (no participant accepts unprofitable tasks)

Asymptotic optimality (as the supply of provers grows, allocation approaches maximum efficiency).

This market runs on the Brevis Chain, which is a dedicated Rollup architecture designed specifically for auction coordination. This architecture isolates market throughput from Layer 1 or Layer 2 network congestion while maintaining transparency and permissionless participation. Proofs generated through ProverNet can be validated against any target blockchain.

Pico zkVM and Real-time Proof

ProverNet builds upon Brevis's existing proof infrastructure, which includes two complementary products to serve different computational needs.

Pico zkVM adopts a "general core + high-performance coprocessor" architecture, where an ultra-efficient core is connected to a dedicated cryptographic accelerator, allowing programs to run a stable virtual machine while benefiting from targeted hardware optimizations.

Pico Prism recently achieved 99.6% proof coverage of an Ethereum block with a gas limit of 45 million, completing 96.8% of the proofs in 12 seconds, with an average proof time of 6.9 seconds. The system uses a 64×RTX 5090 GPU cluster and achieves real-time proofs at 50% lower hardware cost compared to the previous market-leading zkVM.

The ZK Data Coprocessor enables smart contracts to access historical blockchain data and perform off-chain verifiable computations.

Applications supported by this system include PancakeSwap's VIP rate (pre-transaction check of 30 days of trading volume), Euler's trustless reward distribution (processing time-weighted balances of thousands of addresses), and Uniswap v4's route rebates (verification of transaction eligibility without a centralized ledger).

These integrations demonstrate the data-intensive proof requirements that differ from general-purpose computation verification.

Together, these systems establish the economic feasibility of real-time cryptographic verification at the foundational layer while serving the heterogeneous workload types that incentivize the ProverNet market architecture.

Impact on the ZK Ecosystem

ProverNet represents a shift from a single proof supplier's infrastructure to market-driven resource allocation. Existing prover networks are usually optimized for narrow use cases (e.g., specific Rollup proofs, specific virtual machines, or homogeneous work types), limiting their ability to efficiently serve diverse demands.

The market auction model allows proof teams to focus on specific optimization goals (e.g., STARK-based batch processing, low-latency SNARK proofs, specific cryptographic operations) without needing to build generic infrastructure. Applications can then benefit from competitive pricing and service guarantees without being subject to vendor lock-in.

For developers, this eliminates the dilemma of balancing building custom proof infrastructure (expensive and time-consuming) and relying on centralized services (introducing trust assumptions) because cryptographic verification ensures computation correctness while market competition ensures pricing efficiency and resource availability.

More Information

The full ProverNet whitepaper can be found at the following address: http://www.brevis.network/whitepaper/provernet.pdf

A concise technical overview has been published on the Brevis blog: https://blog.brevis.network/2025/10/28/brevis-provernet-building-the-open-marketplace-for-zero-knowledge-proofs/

This article is contributed content and does not represent the views of BlockBeats.