What is TPS in crypto?

Think of the checkouts at a supermarket: how many customers they serve in total per minute is one thing, and how long each person waits in line is another. The distinction between TPS and a transaction’s wait time works much the same way.

To calculate TPS, we divide the transactions the network processed over a period by the number of seconds in that period. If, for example, it processed 600 transactions in 60 seconds, the average is 10 TPS. That describes that specific period, not necessarily the network’s upper limit.

Processing capacity depends on factors such as block capacity, how often blocks are produced and the computational cost of each transaction. A simple token transfer doesn’t necessarily require the same work as executing a complex smart contract.

Why it matters

When the transactions being submitted exceed the available capacity, a queue can build up. Depending on the network’s design, this can mean longer waits, higher fees or failed submission attempts. For anyone sending crypto or using an app, that’s the practical side of TPS.

But be careful: high TPS doesn’t mean instant finality. Finality is about when a transaction is considered irreversible under the network’s rules. A network can process many transactions at once without each one being finalized immediately.

What to watch out for

When you see an impressive TPS figure, we suggest you first look at what exactly it measures. Is it a theoretical maximum, the result of a controlled test or real activity on the live network? Does it include only successful user transactions, or internal protocol messages too?

Comparisons also need common criteria: the same type of transactions, similar conditions and a clear distinction between the base network and additional processing layers, such as Layer 2s. A low observed TPS may simply mean low usage. And a high number on its own doesn’t tell us how secure or decentralized the network is.