OpenGradient (∇, ∇)

We are excited to announce that OpenGradient has raised $8.5M in seed funding to build a groundbreaking end-to-end decentralized platform for secure open-source AI. Big thanks to our backers @a16zcrypto CSX, @cbventures, @svangel, @balajis, @ilblackdragon, @sandeepnailwal and more! We are proud to also have the support of other visionary investors on this exciting journey including: @CanonicalCrypto, @symbolicvc, @SALTConference, @ForesightVen, @NEARProtocol, @CelestiaOrg, @blackdragon_vc, @ThanefieldRes, @PragmaVentures, @ai, @ekrahm, @Bfaviero, @RyanWatkins_, @HighCoinviction, @chainyoda, @mraltantutar, @PaulTaylorVC, @0xlukeskywalker, @no89thkey, @0xnavage. Check out our official blogpost here: opengradient.ai/blog/opengradi…

@ekang426 @NEARWEEK @NVIDIAGTC @ilblackdragon @near_ai Near cooking 🫡

Not many gets invited to speak at @NVIDIAGTC 3 years in a row. So so proud of @ilblackdragon!🥳 Also looking forward to @near_ai happy hour today “near” the GTC venue! 👀

9/ Explore the model on the OpenGradient Model Hub: 🔗: hub.opengradient.ai/models/OpenGra… More models: 🔗: hub.opengradient.ai

8/ Dynamic fee mechanisms are particularly relevant in the context of Uniswap v4, where programmable hooks allow pools to implement custom fee logic. Models like this could enable pools to adjust trading fees in response to changing market conditions. This work builds on funded research conducted with support from the Uniswap–Arbitrum Grant Program (UAGP). 🔗: opengradient.ai/blog/dynamic-a…

OpenGradient Model Highlight: AMM Fee Optimization Model This model explores a volatility-aware approach to AMM fee design. Instead of relying on static trading fees, the mechanism adapts fee levels using short-horizon volatility forecasts. 🧵 👇🏻

@0xchlorinee @memsync_ai perfect!!

@OpenGradient @memsync_ai submitted my feedback

We’re approaching the final stages of Expertly. Designed as a combination of twin.fun and @memsync_ai, it focuses on building AI experts that are context-aware and task-specific by design. As we finalize the product, we’re gathering input to guide what should be built first. Share your feedback: expertly.so/surveys/ai-exp…

@cysic_xyz 🫡🫡

ICYMI, we’ve joined forces with @OpenGradient to bring Cysic’s ZK proving infrastructure into zkML and advance verifiable AI inference!

@Tipwotip 🔥

@OpenGradient how's it?

What if complex tasks could happen automatically? Yes its possible, as @OpenGradient is focusing on making agents, that can handle tasks automatically Want to book a flight? Agents could one day find the best price and book it for you with every step verified & trustless To understand the future of agents in action watch below 👇

@megumidoteth good clip megumi! 🫡

𝐞𝐦𝐩𝐨𝐰𝐞𝐫𝐢𝐧𝐠 𝐃𝐞𝐀𝐈: 𝐜𝐨𝐦𝐩𝐨𝐬𝐚𝐛𝐢𝐥𝐢𝐭𝐲, 𝐮𝐭𝐢𝐥𝐢𝐭𝐲, 𝐢𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 composability transformed defi through permissionless innovation @OpenGradient is bringing that same principle to AI agents by building protocol-level interoperability that enables agents work together trustlessly to build a productive economy 𝐭𝐡𝐢𝐬 𝐜𝐥𝐢𝐩 𝐞𝐱𝐩𝐥𝐚𝐢𝐧𝐬 𝐰𝐡𝐲 𝐚𝐠𝐞𝐧𝐭-𝐭𝐨-𝐚𝐠𝐞𝐧𝐭 𝐢𝐧𝐭𝐞𝐫𝐚𝐜𝐭𝐢𝐨𝐧𝐬 𝐚𝐫𝐞 𝐭𝐡𝐞 𝐤𝐞𝐲 𝐝𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐭𝐢𝐚𝐭𝐨𝐫 𝐛𝐞𝐭𝐰𝐞𝐞𝐧 𝐠𝐞𝐧𝐮𝐢𝐧𝐞 𝐃𝐞𝐀𝐈 𝐢𝐧𝐟𝐫𝐚𝐬𝐭𝐫𝐮𝐜𝐭𝐮𝐫𝐞 𝐚𝐧𝐝 𝐡𝐲𝐩𝐞 most AI platforms operate as closed systems where agents can not interoperate or verify each other's outputs what is opengradient building? it is building the infrastructure to fix that problem with: - x402 protocol integration for native agent payments - TEE + zkML verification for cryptographic proof of inference - protocol-level composability for cross ecosystem coordination - HACA architecture for fast decentralized compute 𝐰𝐡𝐲 𝐭𝐡𝐢𝐬 𝐦𝐚𝐭𝐭𝐞𝐫𝐬: the vision of digital twins that can invest, trade or execute at scale is not just an idea but something that is possible with the right infrastructure like - twin(.)fun: an AI digital marketplace - bitquant: defi analytics agent - model hub: web3 native AI models - memsync: persistent agent memory understand that without verifiable interoperability, agent economies will remain constrained by manual intervention for that, @OpenGradient is building the foundation so that agents can: - negotiate autonomously - execute transactions trustlessly - access specialized ML models - verify outputs cryptographically - coordinate across protocols without permission all this is what separate genuine DeAI infrastructure from hype

@cysic_xyz @cysic_xyz 🤜🤛 @OpenGradient

Cysic is partnering with @OpenGradient! We’re excited to support OpenGradient’s verifiable AI stack by accelerating the zero-knowledge proofs that verify model execution.

More to come as we go deeper on this integration. Read more: docs.opengradient.ai 7/7

This is what the verifiable AI stack actually needs to scale. Not just better models. Not just better chains. The proving layer has to keep up, and now it can. 6/7

We're partnering with @cysic_xyz Together, we're accelerating zero-knowledge proofs that verify model execution on OpenGradient. This is a big step toward making verifiable AI inference production-ready. 1/7