Home Expressions
Docs
Drivers Gateway SDKs Benchmarks
Changelog
GitHub
Blog Status Roadmap
← Back to benchmark index
Historical pipeline snapshot December 23, 2025

December 23, 2025 Pipelining Snapshot

Sequential versus pipelined query throughput under a 50M-query local workload.

172x

pipeline/sequential ratio in this dated run

1.04x
qail-pg / libpq
throughput ratio
1.16x
qail-pg / pgx
throughput ratio
16x
qail-pg / asyncpg
throughput ratio

50M Query Throughput

DriverLanguageQueries/sRelative to qail-pg
qail-pgRust353,0001.00x
libpqC339,0000.96x
pgxGo303,0000.86x
asyncpgPython~21,000~0.06x

Connection Pooling Snapshot

1.25M q/s was recorded in a ten-connection pipelining run in the same dated environment. Treat that number as a pool-and-pipeline configuration result, not as a single universal throughput figure.

Methodology Snapshot

  • Workload target: 50 million query operations with sequential and pipelined variants.
  • Comparison group: qail-pg, libpq, pgx, and asyncpg under comparable local conditions.
  • Primary metric: observed throughput in queries per second plus normalized ratios.

Interpretation Notes

The 172x value reflects a specific sequential-versus-pipeline contrast in this run, not a universal multiplier. Network topology, query shape, server tuning, and pooling settings all shift real-world outcomes. Use the report as directional evidence and validate against your own latency budget and production traffic model.