Solana (SOL) Processes 8.92 Billion Non-Vote Transactions in Q3 2026

The 8.92 Billion Quarter

Solana (SOL) closed the third quarter of 2026 with its heaviest usage on record, processing 8.92 billion non-vote transactions according to figures the project's official X account published on Monday. The total runs 104% ahead of the roughly 4.38 billion non-vote transactions the network handled in Q4 2025, and it is the first quarter in the available sequence to clear the 8 billion mark. The quarterly path carries as much weight as the headline: activity reached 7.54 billion in Q1 2026, slipped to 7.19 billion in Q2, then climbed about 24% quarter-over-quarter to the 8.92 billion print, undoing the second quarter's dip. The metric excludes the routine ballots validators emit to keep the chain's consensus mechanism running, which is why it is treated as a closer proxy for real demand than raw totals. Validators, the operators who post staking collateral to secure the chain, send those votes continuously, and stripping them out leaves the economic traffic: token transfers, dapp usage, trading and other user movement on-chain. Solana (SOL), a layer-1 blockchain built for high throughput and low fees, has leaned on that capacity to absorb bursts of application load without the congestion that has strained smaller chains. The milestone lands with the Solana price near $104 on Oct. 19 and a market capitalization of about $61 billion, and the tape around that level sits on our Solana technical analysis page. Activity of this scale across the Solana ecosystem is what the throughput was built for.

Yakovenko's Six-Word Reply

Solana co-founder Anatoly Yakovenko, widely known as Toly, answered the quarterly record in characteristic fashion. His post on X reads, “Put the cryptography into the computer,” a nod to the engineering behind the throughput numbers rather than a remark on markets. The monthly breakdown underneath the quarterly total is where the detail sits. Data compiled by Solana (SOL) Compass puts completed non-vote transactions at roughly 3.18 billion for September alone, the third consecutive monthly record, a pace consistent with the 8.92 billion booked across the full quarter. September's 3.18 billion implies the two prior months also set records, with July and August each clearing their own previous highs, a cadence that matches the acceleration visible in the quarterly figures. The standing caveat is bot traffic. Automated traders generate a large slice of on-chain volume, so transaction counts do not convert one-for-one into human users, and a single high-frequency account can push thousands of transactions an hour. The same caution applies to the quarterly record itself: 8.92 billion transactions says the chain was busy, not that 8.92 billion people did anything, and the distance between those two statements is where most of the debate over Solana usage numbers lives. That distortion is why app-level engagement gets tracked separately, and our rundown of the top Solana DApps maps where user-facing activity actually concentrates. None of the figures published on Monday carry a user-count breakdown, so the split between bot and human activity remains undisclosed. Yakovenko's line is an engineer's framing of the same point: the milestone he chose to celebrate was computational, not financial.

@toly · X post

His post on X.

View on X

Bots, Fees and the Alpenglow Test

Two threads come out of Monday's data. The 104% jump confirms that throughput demand is real, but volume without revenue tells only half the story, and fee income is where inflated traffic gets priced out; our earlier review of fee dominance on the network showed how sharply that share can swing between quarters. The second thread is timing. The Alpenglow upgrade targets October 21 activation, days after the record quarter closed, which hands the network a clean before-and-after benchmark for a major protocol change. If post-upgrade volumes hold at or above the Q3 pace, the record reads as genuine adoption; if they fall back, the bot share was doing more lifting than the headline admitted. Either outcome will show up in fees before it shows up anywhere else.