Regulated digital-market venue 21X AG and LayerZero have reported a high-throughput test in which 21X's central limit order book ran on a private, permissioned deployment of the Zero blockchain. The partners say the system processed 10,000 order matches per second in their month-long exercise, with matching and settlement completed within the same transaction. The finding is a technical benchmark, not an exchange-volume figure from the public market.[1]
The comparison is against common EVM deployments used in the companies' test framework. The release describes an improvement of up to 45 times against a selected EVM layer-two baseline. That ratio implies a benchmark comparator of about 222 matches per second if both numerator and denominator represent the same workload. It does not establish that every competing blockchain performs at 222 matches per second, or that a public Zero deployment already sustains the higher figure under normal market conditions.[1]
How the benchmark should be read
The distinction between an onchain and offchain order book matters. A conventional trading venue typically matches orders in a privately operated engine and sends information or settlement instructions elsewhere. In the pilot, 21X says the central limit order book itself matched orders on the network. The test therefore targets a known architectural bottleneck, not simply a faster way to publish trade confirmations after decisions have been made offchain.[1]
A permissioned deployment is relevant because institutions may need to restrict node access and transaction visibility. The parties describe Zero as EVM-compatible, offering a route to existing development tools while controlling who participates. But permissioning changes the operating environment; it can produce a performance profile different from an open network with unrelated participants and different consensus, connectivity and monitoring conditions.[1]
Evidence and operating implications
| Measure | Finding or scope |
|---|---|
| Participants | 21X AG and LayerZero |
| Network | Zero, private permissioned deployment |
| Test duration | About one month |
| Reported matching rate | 10,000 matches/second |
| Reported speed advantage | Up to 45x versus selected EVM L2 baseline |
| Implied baseline at 45x | Approximately 222 matches/second |
| Commercial status | Pilot; separate 21X EU system is already live |
The production questions still unanswered
21X is already an authorised European DLT trading and settlement venue, with production infrastructure separate from this experiment. The company explicitly says that its EU system continues operating alongside the Zero work. Its US affiliate, 21X LLC, also warns that it is not a registered US broker-dealer. A licence or successful performance test in one jurisdiction does not authorise an equivalent commercial launch in another.[1]
The commercial test will require repeatable, independently inspectable benchmarks: message sizes, order cancellations, peak bursts, latency distributions, data-retention costs and recovery after interruption. These are the numbers brokers and market makers need before relying on a headline matching speed. The November release provides a useful upper-throughput claim under stated conditions; it does not yet supply those full operating metrics.[1]
Sources & further reading
- 21X AG and LayerZero deliver institutional-grade onchain trading pilot on private-permissioned deployment of Zero 21X AG / LayerZero · 4 Nov 2026 · Accessed 7 Nov 2026