Overview
Gevulot is building a decentralized proving layer — a shared infrastructure network where any application can submit proof generation requests, and a permissionless set of provers competes to generate them. The key insight is that ZK proving is computationally expensive and most projects can't afford to run dedicated proving infrastructure. By creating a shared marketplace, Gevulot enables cost sharing across many applications while providing provers with better hardware utilization.
The protocol is prover-system agnostic — it can generate proofs for any ZK framework (SP1, RISC Zero, Plonk, Groth16, STARKs, etc.). This generality makes Gevulot complementary to specific zkVMs rather than competitive. SP1, RISC Zero, and others define what gets proved; Gevulot provides the infrastructure for where and how proofs are generated.
The architecture uses a custom blockchain for coordination (proof request submission, prover selection, payment settlement) while actual proving happens off-chain on prover hardware. The economic model is straightforward: applications pay for proofs, provers earn revenue, and the protocol facilitates the marketplace.
Technology
Gevulot's proving layer coordinates proof generation across a distributed network of prover nodes. The protocol handles workload scheduling, prover selection (based on capability and performance), proof verification, and payment settlement. The system supports heterogeneous prover hardware (GPUs, FPGAs, custom accelerators) and multiple proving frameworks.
The key technical challenges are efficient proof request routing (matching requests to capable provers), latency optimization (minimizing time from request to proof delivery), and verification (ensuring submitted proofs are valid). The protocol uses a custom consensus mechanism for coordination and supports batch proving for cost efficiency.
Security
The security model ensures provers can't submit invalid proofs (verification on the coordination layer) and applications receive correct proofs within guaranteed timeframes. Economic incentives (staking, slashing) align prover behavior with honest computation. The protocol's security is critical because invalid proofs could compromise the applications relying on them. The coordination layer itself must be secure against manipulation.
Decentralization
Gevulot is designed for permissionless prover participation — anyone with capable hardware can join the network. The marketplace structure prevents monopolization by distributing work across competing provers. Geographic and hardware diversity improves network resilience. The governance model aims for progressive decentralization as the protocol matures.
Adoption
Adoption is early-stage with partnerships across the ZK ecosystem. Rollups, bridges, and ZK-application developers are the target customers. The value proposition (cheaper, more reliable proof generation) is compelling, but integrating with a new proving network requires trust and technical effort. Early deployments are proving the concept.
Tokenomics
The token facilitates the marketplace economy — applications pay in tokens, provers earn tokens, and staking aligns incentives. The economic model captures value from the growing ZK proof market. As more applications require proofs, the demand for proving services (and tokens) should grow. The tokenomics are well-structured around a clear utility: paying for computational work.
Risk Factors
- Early stage: Network and technology are still developing
- Adoption timing: ZK proof demand must grow to justify the infrastructure
- Competition: Succinct's proving network, Aligned Layer, and others target similar markets
- Hardware centralization: Efficient proving requires expensive hardware
- Coordination complexity: Distributed proving introduces latency and reliability challenges
- Market size uncertainty: The total addressable market for proving services is unclear
Conclusion
Gevulot addresses a real infrastructure gap — shared proving services can reduce costs and complexity for the growing ZK ecosystem. The prover-system-agnostic approach ensures broad compatibility, and the marketplace model creates efficient resource allocation. The project is well-positioned in the ZK infrastructure stack as a complementary service to zkVMs. Success depends on the overall growth of ZK applications and Gevulot's ability to build a reliable, efficient proving network. The concept is sound and the market need is real.