A payment-channel design on Solana sustained throughput above 1 million transactions per second in a test run across 100,000 wallets, a figure the network reported Wednesday that dwarfs its typical onchain pace.
The architecture lets an agent authorize a spending limit once, process multiple payments offchain, then settle the final balance in a single onchain transaction — a structure that trims per-payment cost and latency versus recording each transfer on the ledger. That compression matters for microtransactions, where network fees can exceed the value of a single payment.
The result compares with Solana's theoretical throughput of up to 65,000 transactions per second and real-world speeds of roughly 1,100 to 6,200 TPS, per the network's published specifications. Solana already carries more than $3 billion in tokenized real-world assets, and seven global systemically important banks, including Morgan Stanley and JPMorgan Chase, are building on the chain, according to the Solana Foundation.
The test's significance rests on whether it moves beyond the lab. Solana's offchain-plus-settlement model targets a fast-growing niche: payments initiated by AI agents, which can fire thousands of microtransactions a second. Payments infrastructure has become the limiting factor for AI-driven software, as companies ship features faster than legacy settlement rails can adapt, and crypto rails are competing to fill that gap.
Named protocol deployments on the mainnet, not a benchmark figure, will determine whether the milestone lifts SOL and pulls developers toward Solana-based payment rails over competing settlement layers such as Ethereum. For now the announcement is a narrative win for a network already courting institutions: Morgan Stanley has built two Solana-based products, including the Morgan Stanley Solana Trust, while JPMorgan has issued on-chain commercial paper on the chain.
Whether the throughput claim translates into live, audited payment volume — and which protocols adopt the channel design — is the metric traders will watch next.
This article is for informational purposes only and does not constitute investment advice.