Optimized Zebra
6.8 ms
Median of 8 blocks
Latest block 3,536,350: 1.9 ms
Updated 2026-10-30T23:10:30Z
Every new Mainnet block goes to all three nodes at once. The numbers are how long each node takes to accept it, over the last 24 hours. Optimized Zebra: zebrad 7.0.0+227.g5527c56, pasta-asm
15.9× faster
Optimized Zebra vs Zebra v7.0.0
Saves 115 ms per block (median of 8 blocks)
1.08× faster
Optimized Zebra vs Zakura
Saves 0.49 ms per block (median of 8 blocks)
6.8 ms
Median of 8 blocks
Latest block 3,536,350: 1.9 ms
122.0 ms
Median of 8 blocks
Latest block 3,536,350: 9.3 ms
7.5 ms
Median of 8 blocks
Latest block 3,536,350: 1.9 ms
The chart needs JavaScript.
| Size | Blocks | Zakura | Zebra v7.0.0 | Optimized Zebra | Optimized vs Zebra v7.0.0 | Optimized vs Zakura |
|---|---|---|---|---|---|---|
| All sizes | 8 | 7.5 | 122.0 | 6.8 | 15.9× faster | 1.08× faster |
| <20 kB | 3 | 2.6 | 10.9 | 2.3 | 4.97× faster | within 5% (1.04×) |
| 20–300 kB | 5 | 7.8 | 122.0 | 7.1 | 18.1× faster | 1.09× faster |
| >300 kB | 0 | — | — | — | — | — |
| Height | kB | Zakura | Zebra v7.0.0 | Optimized Zebra |
|---|---|---|---|---|
| 3,536,350 | 1.630 | 1.9 | 9.3 | 1.9 |
| 3,536,349 | 33.352 | 7.0 | 119.2 | 6.6 |
| 3,536,348 | 2.381 | 2.6 | 10.9 | 2.3 |
| 3,536,347 | 47.162 | 8.2 | 123.0 | 8.4 |
| 3,536,346 | 29.117 | 7.8 | 121.9 | 7.1 |
| 3,536,345 | 13.531 | 16.4 | 169.2 | 15.8 |
| 3,536,344 | 23.387 | 7.2 | 122.0 | 6.1 |
| 3,536,343 | 27.119 | 9.5 | 721.1 | 8.4 |
A Zebra node follows Mainnet and hands each new block to the three nodes with submitblock. The three nodes have no peers and empty mempools, so the time is mostly checking and storing the block. They share one machine with other work. Zakura keeps a pruned chain; both Zebra nodes keep the full chain.