Here we report on the progress of the leading builders in the Coprocessor ecosystem, documenting recent significant releases, technical breakthroughs and general update
@lagrangedev has launched DeepProve, a zkML library that makes AI inferences verifiable and up to 158 times faster than existing tools. As AI grows more powerful and opaque, DeepProve enables proof that AI outputs are correct and aligned with human intent. By combining ZKPs with machine learning, Lagrange aims to ensure AI systems act safely and transparently, helping prevent harmful or deceptive outcomes: https://lagrange.dev/blog/announcing-deepprove-zkml…
They also announced that @the_matter_labs will route up a significant portion of its outsourced proof workload to Lagrange’s Prover Network over the next two years. This follows successful integration with @zksync’s ZK stack and highlights progress toward decentralizing proofs. The LPN improves cost, speed, and reliability, supporting developers building sovereign ZK chains. It also boosts performance for users and businesses on ZKsync’s Elastic Network by reducing gas fees and enhancing scalability: https://lagrange.dev/blog/matter-labs-uses-lagrange-decentralize-proving…
In other news, @ana_heygrowth of Lagrange teased about a partnership with @HouseofZK, with the two projects hosting the ZK/AI Summit on April 28th at @token2049 in Dubai: https://lu.ma/2qhbj7fn The event will explore the intersection of ZK and AI, featuring discussions on verifiable computation, privacy-preserving ML, and scalable systems. Experts from around the globe will examine how ZK can bring trust and integrity to intelligent apps: https://x.com/ana_heygrowth/status/1907406100804305373…
They also announced numerous other partnerships, including:
• @inference_labs: Collaborating to deploy verifiable AI models using Lagrange’s DeepProve zkML, ensuring performance and security compliance without revealing model details: https://x.com/lagrangedev/status/1900231840583344339…
• @QuillAI_Network: Integrating DeepProve with QuillShield and QuillCheck to verify AI-generated security reports in Web3, raising the standard for smart contract auditing: https://x.com/lagrangedev/status/1900583960838606881…
• @OpenledgerHQ: Bringing trust and transparency to AI-powered DeFi trading through verifiable predictions and smart contracts backed by zkML: https://x.com/lagrangedev/status/1901635403783782712
• @UngateAI: Introducing verifiable AI in healthcare using federated learning and DeepProve to ensure diagnostic accuracy and patient privacy: https://x.com/lagrangedev/status/1901997348915474692…
• @hetu_protocol: Enabling verifiable tracking of AI research contributions, protecting data privacy while ensuring scientific recognition and attribution: https://x.com/lagrangedev/status/1902366558459949081
• @0xPredicate: Powering a censorship-resistant policy infrastructure with zkML-verifiable transaction rules, enabling more expressive and secure dApps: https://x.com/lagrangedev/status/1904171921085776191…
• @IQAICOM: Verifying voting and compliance in AI agent operations within DeFi, increasing transparency and trust in agent tokenization: https://x.com/lagrangedev/status/1904533820168093980…
• @peri_labs: Scaling verifiable AI in DePIN environments, allowing edge devices to securely use AI with tokenized compute and model access: https://x.com/lagrangedev/status/1905273641559847343…
• @tempest_fi: Verifying AI-generated liquidity management strategies via the Tempest Terminal’s LLM, ensuring offchain inferences are correct and scalable: https://x.com/lagrangedev/status/1906713130379460975…
• @zerufinance: Enabling trustless, verifiable onchain reputation scores with DeepProve, ensuring that behavior-based reputation (zScore) is calculated fairly and transparently: https://x.com/lagrangedev/status/1907070548070293989
@primus_labs has partnered with @inter_link, a decentralized human network, to improve AI training using verified human data: https://x.com/primus_labs/status/1900514924381777994…
The collaboration combines Primus’s zkTLS technology with InterLink’s AI systems to strengthen data privacy and verification.
In a new episode of @HouseofZK Radio, @xxiang_xie, Co-founder of Primus, joined @alicelingl to talk about the full data lifecycle - from zkTLS for access control to ZK-FHE for private computation. He outlined Primus’s Quicksilver proofs, stressed client-side proving, and explored ZK+FHE use cases in identity, agents, and data monetization:https://x.com/HouseofZK/status/1907060744111231251
@AutomataNetwork published its Monthly Update: https://blog.ata.network/automatas-monthly-update-issue-79-71ab7bab0a62…
In summary, the update includes:
- UI/UX updates to the @1rpc_ and @l2faucet websites
- Preparations for the drand Mainnet Ceremony
- Support for 250ms Flashblocks on @base Sepolia via 1RPC
- Release of the DCAP plugin for @elizaOS running in SGX/TDX- Integration of OP @SuccinctLabs Lite with ZK Fraud Proofs for finality
- Team participation in @EthereumDenver and @consensus_hk on TEEs and AI agents
@vlayer_xyz has released updates to its Testnet, introducing new tools and improvements: https://vlayer.xyz/blog/time-to-test-introducing-vlayers-testnet…
Developers now have access to Web and Email Notary, Call Prover, a local Devnet environment, and integrations with @ethereum, @base Sepolia, and @Optimism Sepolia networks. The team invites developers to test features, provide feedback, and help prepare the ecosystem for the upcoming Mainnet launch.
@HerodotusDev introduced Sentinel, a tool that lets developers enforce custom rules for blockchain transactions using cryptographic accumulators and ZKPs: https://herodotusdev.medium.com/sentinel-pre-checks-practical-rule-enforcement-on-permissionless-blockchains-d57420edd3b6…
Embedded in contracts or at the protocol level, it checks conditions like wallet age, compliance status, or blacklist membership before allowing transactions. It balances blockchain openness with security and compliance needs, enabling smart contracts to verify large or off-chain data sets without revealing sensitive details.
@SpaceandTimeDB is integrating its SXT solution with @heurist_ai to give AI agents access to verifiable blockchain data using ZKPs. This allows agents on the Heurist Mesh to query secure, cross-chain data, enabling trusted automation in trading, governance, and DeFi strategy optimization based on accurate, tamper-proof analytics: https://x.com/SpaceandTimeDB/status/1905661323276681451…
In addition, the projects' ZK-secured blockchain data now powers @heurist_ai’s decentralized AI cloud, enabling verifiable cross-chain analysis and trustless intelligence. By integrating SQL queries with petabytes of ZK-proven data, Mesh agents gain the ability to detect patterns and extract insights across networks: https://x.com/SpaceandTimeDB/status/1905661323276681451
@no89thkey, Founder of @brevis_zk, will participate in The Boundless House of ZK during @festival_web3, co-hosted by @boundless_xyz (@RiscZero) and @HouseofZK: https://x.com/HouseofZK/status/1906974795046011133…Michael will join a panel titled 'A Fight for Mindshare', which will explore the emerging ZK narrative across ecosystems, focusing on developer engagement, user expectations, and educational strategies within the ZK space. Joining him in the discussion will be @sinka2022, Founder of @DelphinusLab, @yq_acc, Founder of @alt_layer, @leofanxiong, Co-founder of @cysic_xyz, and Tina Liu, Head of Global Marketing at @RiscZero
Brevis will also host their own event on the 5th - ZKConnect - featuring leading speakers from @OpenBuildxyz, @alt_layer, @EthStorage, @ProjectZKM and more.
Register today to secure your spot: https://lu.ma/z75uwkls
