Reddit user builds on-chain arbitrage decision layer with Jev, leaving execution to an LLM
A Reddit user connected Jev to on-chain data streams to spot arbitrage opportunities for free, then hands the good ones to an LLM to buy, bridge and sell, at roughly the cost of a basic subscription.
Jev + Opus 5.5 or GPT 6 use case everyone is silent about: and they are right Some random guy on Reddit showed the MVP of what he has created combining Jev with LLM model → First, he explained his AI agent what you can do onchain and where: swaps, bridges, wallet tracking, arbitrage. → Second, he got access to on chain data using public RPCs, so now he has every transaction happening on Top 5 chains → Third, he implemented Jev as a decision maker model which can work with this data almost for free The script saves millions of transactions -> Jev scans them live and decides if there is an arbitrage opportunity -> sends the good options to LLM -> LLM does the rest (buys, bridges, sells) Every micro spread you couldn't see with your eyes is being tracked and closed, which brings him money Costs? Same as you spend on a basic subscription
This post quotes another post: x.com/kaiandai/status/2103871840066130241