Axelar announced AgentFlux, an open source framework designed to run AI agents locally while keeping private keys, trading strategies, and customer data out of the cloud. This is a pitch aimed at institutions considering on-chain finance but wary of privacy risks.
AgentFlux, developed by Interop Labs, the team behind the Axelar network, enables financial firms to deploy “agent” automation without sending sensitive information to external infrastructure, the company announced Thursday.
This framework tackles one of the biggest frictions in AI-driven crypto operations: tool invocation.
Currently, most agents rely on cloud-based models to decide which blockchain tools to call and how to structure transactions. This can inadvertently expose the very information that institutions are trying to protect. AgentFlux splits these tasks into two small specialized models. one for selecting the appropriate tool and the other for generating the arguments to execute. According to the team behind Axelar, this setup improves tool call accuracy by 46% in benchmark tests, bringing local models closer to the performance of large cloud systems.
Axelar co-founder Sergey Gorbunov told CoinDesk in an interview that he sees AgentFlux benefiting multiple tasks. “First of all, they are very sophisticated trading strategists, and they are incredibly unique.” “You never want to get a public cloud model as effective as executing your own strategy, and then when you upload it, you’re essentially violating that model. That’s one of those things.”
“The second thing I’ve seen is that blockchain is also often used for analytics, for example, doing tax filings or investigating who performed which transactions. Or maybe you have a specific fingerprint and you’re anticipating malicious activity across several accounts. We want to ask them to upload their fingerprints to a cloud model or an AI system to understand all the correlated transactions, all the suspicious behavior and patterns that this account has, so these types of activities can be completely stored and done privately,” Gorbunov added.
AgentFlux is also part of Axelar’s broader multi-chain strategy. The team positions Axelar as a “gateway to on-chain finance,” providing the infrastructure for institutions to move assets and data across blockchains from a single integration point. AgentFlux enables a single AI agent to see risk, assess exposure, and trade across multiple ecosystems. The company believes this is an essential feature for businesses to implement.
Read more: Canary Capital files to launch ETF tracking cross-chain protocol Axelar
