Skip to content

Async SQLite Benchmark

Apple-to-apple performance comparison between zerodep aiosqlite and aiosqlite.

Test Environment

  • CPU: x86_64 Linux
  • Python: 3.12
  • Tool: pytest-benchmark 5.2.3 (mean values reported)
  • Reference: aiosqlite 0.22.0
  • Last Updated: 2026-09-28

Implementations

Implementation File/Package Description
zerodep aiosqlite.py stdlib-only async sqlite3 wrapper
aiosqlite (reference) Popular async sqlite3 library

Tests Performed

Test Description
Single Insert+Select Connect, CREATE, INSERT 1 row, SELECT, close
Bulk Insert executemany with 1000 rows
Bulk Select execute_fetchall over 1000 rows
Cursor Iteration async for over 1000-row cursor
execute_insert 100 sequential execute_insert calls
Large Row Insert executemany with 500 rows × 4KB payload
Large Row Select execute_fetchall over 500 rows × 4KB payload
Large Result Fetchall execute_fetchall over 10K rows
Large Result Iteration async for over 10K-row cursor
Concurrent Tasks 8 coroutines, each inserting 50 rows + reading all
Mixed Read/Write 200 interleaved inserts and selects

Run It Yourself

pip install pytest pytest-benchmark aiosqlite
pytest aiosqlite/test_aiosqlite_benchmark.py --benchmark-only -v

Latest CI Results

Updated automatically on each release via Benchmark CI.