Solana Performance Statistics: Live TPS, Compute + Fee Tracker
Transaction Performance
Raw Transactions Per Second
Transactions per second is a measure of the raw throughput of the Solana network. It is helpful to see the health of the cluster as a whole, but doesn't tell the whole picture.
On the one hand, TPS is generally much lower than Solana's theoretical maximum, as the number of transactions submitted each second varies based on how busy the network is. On the other, the raw TPS figure is inflated by validator's own vote transactions, hence the other metrics we have included here
Votes vs Non Votes
Use this chart to see the breakdown of votes vs all other transactions on the Solana network
Success Rate
This is the percentage of transactions that succeeded for each minute in the last hour. Transactions can fail for many reasons, for example due to insufficient funds, excess slippage on swaps or other user errors.
Non Vote Transactions Per Second
Forget the validators - just how many transactions is Solana processing per second, outside of votes? This metric lets you see at a glance how many transactions Solana's end users are using.
Vote Transactions Per Second
How many votes are being processed by the Solana network per second? Since every validator votes on every slot, this figure should scale with greater number of validators and lower block times.
Blocks Per Second
How many blocks are being processed by the Solana network per second? The expected time for each block to be processed is 400ms. At times of high congestion this can rise considerably. Your staking rewards are directly affected by block times: the shorter the duration, the higher the APY you receive on your stake.
Block Time
How long does it take for a block to be processed by the Solana network? The expected time for each block to be processed is 400ms. At times of high congestion this can rise considerably. Your staking rewards are directly affected by block times: the shorter the duration, the higher the APY you receive on your stake.
Transactions Per Block (TPB)
Unlike other metrics in this set, to increase readability Transactions Per Block shows the last 750 blocks, which is roughly the last 5 minute period. Gaps in the chart represent slots skipped by validators.
Historic Performance Stats
Seen an interesting bump on the chart? Dig in using our block explorer to see the last 60 minutes's blocks
Live Fee data
Fee to Compute Unit Ratio
Solana transactions are ordered by validators based on how much compute they will consume, vs how high the fee is. This means smaller, less intensive tx can land for much lower fees than big, heavy, slower transactions.
The median fee shows how many lamports (1 billionth of a SOL) it costs per compute unit for the majority of transactions, the average can be skewed by a handful of expensive transactions.
Average User Priority Fees
When looking at averages a single transaction with a high fee can skew the mean fee dramatically, even if thousands of user transactions paid far less.
For an accurate comparison with other blockchains it is best to look at the median fee paid across all user transactions. Note this does not include fees from vote transactions which would unfairly lower the median.
Max User Priority Fees
What is the maximum fee users are willing to pay to have their transactions processed? This chart shows the highest and lowest fees paid by users to have their transactions processed in the last 60 minutes.
The highest fees tend to be paid to access the 'hottest' accounts - for example the trading pools with the highest volume - and are not indicative of the usual fees paid by users elsewhere on the network.
Historic Fee Data
Want to see how much is spent each day in fees, or how much SOL is burned? Check out our new Fee Tracker!
Compute Performance
Compute Units Per Second
While Transactions Per Second (TPS) has become a popular metric to measure Solana's performance, it doesn't tell the whole story.
Low TPS may mean the network is underperforming - or it simply may be a quiet time. Conversely it may be that the types of transaction - and the programs they invoke - are more demanding than average.
Looking at Compute Units helps us see how hard the validators are working, not just how fast. It also excludes votes by default - all recorded compute units come from user activity on Solana.
Success vs Failed Compute Units Per Second
Another way to look at Compute Units is to understand how much compute power was used by validators to process failed vs successful transactions.
Blockspace Usage: Percent of Max Compute Used
Each block processed by Solana has a maximum of 48,000,000 compute units available. This chart shows the percentage of that total that was used by validators to process transactions, expressed as an average for each minute of the last hour.