Here we report on the progress of the leading builders in the Proving Service ecosystem, documenting recent significant releases, technical breakthroughs and general updates.
Featuring: @brevis_zk, @cysic_xyz, @fermah_xyz, @MarlinProtocol, @vlayer_xyz, @zan_team, & @thezkcloud.
ProverNet
@brevis_zk has launched its ProverNet mainnet beta, turning its newly published proof-market whitepaper into a functioning decentralized marketplace: https://blog.brevis.network/2025/12/08/brevis-provernet-mainnet-beta-is-live/
Applications can now submit proof requests without running their own infrastructure, while provers compete in continuous auctions with real-time payment clearing. USDC will be replaced by BREV at full launch.
Early workloads, including Ethereum block execution proving, are being migrated as ProverNet moves toward dedicated rollup deployment.
Events
Brevis took the stage at Verifying Intelligence 3.0 organized by @HouseofZK during @EFDevcon, presenting one of the event’s core keynotes: https://x.com/HouseofZK/status/1996580816923586810
Co-founder & CEO @no89thkey outlined how verifiable computation shifts blockchains from executing all workloads to efficiently verifying them. He explained how this model enables data-intensive DeFi logic, historical on-chain analytics, and effectively “infinite compute” for Web3 without exceeding gas limits.
Media
Finally, Michael joined @alicelingl on a recent episode of House of ZK Radio: https://x.com/HouseofZK/status/1993675089162699254
They talked about Brevis’ path from its DeFi origins to today’s focus on verifiable compute and near real-time proving.
Michael explained the Pico zkVM and Pico Prism distributed prover, showing how their modular glue-and-coprocessor model supports loyalty systems, reward distribution, and CEX/DEX bridges. He also outlined how developers can start building with Brevis or join the Proving Ground campaign.
Full podcast: https://hozk.io/radio#91-michael-co-founder-of-brevis
Media
In an article by @blockchainrptr, @cysic_xyz Founder @leofanxiong, shared how academic research shaped his decision to build dedicated ZK hardware after seeing real-world systems constrained by slow, inefficient proving: https://blockchainreporter.net/from-cornell-lab-to-crypto-protocol-dr-leo-fan-on-building-cysics-zk-engine/
He explains why Cysic’s silicon-to-protocol model, custom accelerators, and decentralised prover network are central to scaling verifiable computation.
Leo outlines roadmap shifts toward AI and ZK workloads, incentive design for operators, and a long-term vision of Cysic as core infrastructure for verifiable compute.
Partnerships
Cysic announced a partnership with @NetworkNoya to accelerate the proving layer behind Noya’s zkML-based AI agents for multi-chain liquidity management: https://x.com/cysic_xyz/status/1991872401014137001
Noya uses ONNX models routed through @ezklxyz and Halo2/KZG for EVM verification on @base, where proving has become the heaviest stage.
By offloading workloads to Cysic’s optimized backend, Noya gains lower proving costs, reduced latency, and higher throughput without compromising trustlessness.
@fermah_xyz Founder @vanishree_rao kicked off the second series of @HouseofZK's "ZK Industry Leaders", diving into the intricacies of Proof Markets: https://youtube.com/watch?v=KCdYFgK38cU?t=1
In the episode, Vanishree outlines Fermah’s approach to unified proving infrastructure, covering pricing, orchestration, upgradeability, and predictable costs, and examines server - versus client-side proving, confidential delegation, and the trade-offs shaping future decentralized proof markets.
@MarlinProtocol and @BinanceAcademy introduced a free course on offchain computing with TEE coprocessors, aiming to give developers practical skills in confidential computing: https://blog.marlin.org/binance-academy-and-marlin-launch-public-course-on-tee-coprocessors
The program goes deep on why decentralized systems depend on offchain execution and walks through TEE architecture, deployment patterns, and real AI and DeFi applications.
Through hands-on modules led by a Marlin contributor, participants learn how TEEs provide secure, scalable performance for advanced dApps.
@vlayer_xyz demonstrated how its proof infrastructure can automate conditional payments in global trade through a new collaboration with @Damisaxyz: https://vlayer.xyz/blog/damisa-and-vlayer-pioneer-a-new-era-of-automated-proof-driven-payments-in-global-trade-1
The teams tested a system that releases escrowed funds only after a verified delivery confirmation. vlayer authenticated the email signal, extracted the transaction ID, generated a verifiable proof, and triggered instant settlement within Damisa’s payment rails.
The result shows how validated real-world events can drive automated, trusted payment flows across cross-border supply-chain operations.
Ethproofs Day
@thezkcloud shared an article covering an @eth_proofs Day demo during @EFDevcon, where @drakefjustin used a zkLighthouse client to validate blocks solely through proof verification: https://blog.zkcloud.com/p/a-glimpse-into-ethereums-zk-future
The event marked a concrete step in @ethereum’s shift from re-execution toward proof-based validation. ZkCloud, alongside five zkVM teams, delivered real-time proofs for mainnet blocks, with @ziskvm meeting all criteria for real-time performance.
The demo showed how multiple independent proofs allowed attestations without an execution client, offering an early view of Ethereum’s emerging ZK-first direction.
Verifying Intelligence
ZkCloud also took part in Verifying Intelligence 3.0 organized by @HouseofZK during Devconnect, contributing to one of the event’s key discussions: https://x.com/HouseofZK/status/1993629205116887092
@NorbertVadas, Head of Product at ZkCloud, joined host @alicelingl, @GeoffTRichard of @OntologyNetwork, @jgorzny of @ZircuitL2, and @o1coby of @o1_labs & @MinaProtocol to examine how AI is reshaping governance, identity, and autonomy, and why verifiable computation is becoming essential for maintaining public agency.
