GitHub Repository

A SIMD boosted high-performance and correct Python JSON parsing library, faster than the fastest.

37 starsC

SsrJSON: faster than the fastest Python JSON library

Name: SsrJSON: faster than the fastest Python JSON library
Availability: InStock
Author: antares0982

by antares0982·Feb 24, 2026·4 points·0 comments

Visit Project View on HN

AI Analysis

●●●BangerWizardryBig Brain

Beats orjson via UTF-8 cache exploitation and SIMD; actual speed wins, real benchmarks.

Strengths

•Addresses orjson's UTF-8 cache blind spot with measurable gains on non-ASCII data (2-3x on Chinese strings).
•SIMD + yyjson codebase is genuinely fast; author blogs detailed technical approach, not marketing hype.
•Drop-in stdlib replacement eliminates adoption friction; no API rewrite needed.

Weaknesses

•Niche win: most JSON workloads are I/O-bound, not CPU-bound; only benefits high-throughput pipelines.
•Early maturity (v0.0.9); orjson has years of battle-testing and wider ecosystem trust.

Post Description

I developed a Python JSON library, ssrJSON, built on the yyjson codebase and SIMD techniques. It can serve as a compatible replacement for most common JSON use cases, and benchmarks show it outperforms orjson, which advertises itself as the fastest. Compared with other Python JSON libraries, ssrJSON addresses UTF-8 cache related performance pitfalls, giving it a significant advantage when handling strings; technical details are available on my blog: https://en.chr.fan/2026/01/07/python-json/. In addition, for floating‑point handling ssrJSON uses the latest zmij algorithm.

I noticed that orjson has closed its issue tracker: https://news.ycombinator.com/item?id=47022606. If the community wishes to maintain a stable, high‑performance JSON parser, ssrJSON could be considered as an alternative.