#simd #json #parser #performance


High performance JSON parser based on a port of simdjson

19 releases

✓ Uses Rust 2018 edition

new 0.1.19 Sep 18, 2019
0.1.18 Sep 4, 2019
0.1.17 Jul 29, 2019
0.1.16 Jun 29, 2019
0.1.5 Apr 29, 2019
Download history 117/week @ 2019-06-05 196/week @ 2019-06-12 192/week @ 2019-06-19 302/week @ 2019-06-26 179/week @ 2019-07-03 91/week @ 2019-07-10 119/week @ 2019-07-17 130/week @ 2019-07-24 190/week @ 2019-07-31 185/week @ 2019-08-07 149/week @ 2019-08-14 197/week @ 2019-08-21 135/week @ 2019-08-28 190/week @ 2019-09-04 184/week @ 2019-09-11

742 downloads per month



SIMD Json for Rust   Build Status Windows Build Status Latest Version

Rust port of extremely fast simdjson JSON parser with serde compatibility.

readme (for real!)

CPU target

To be able to take advantage of simdjson your system needs to be SIMD compatible. This means to compile with native cpu support and the given features. Look at The cargo config in this repository to get an example.

simd-json.rs supports AVX2, SSE4.2 and NEON.


If you are writing performance centric code, make sure to use jemalloc and not the system allocator (which has now become default in rust), it gives a very noticeable boost in performance.


simdjson-rs is compatible with serde and serde-json. The Value types provided implement serializers and deserializers. In addition to that simdjson-rs implements the Deserializer for the parser so it can deserialize anything that implements the serde Deserialize trait.

That said serde is contained in the serde_impl feature which is part of the default feature set, but it can be disabled.


simdjson-rs is not capable of serializing JSON data as there would be very little gain by re-implementing it. For serialization, we recommend serde-json.


simdjson-rs uses a lot of unsafe code first of all since all SIMD-intrinsics are inherently unsafe and also to work around some bottlenecks introduced by rust's safe nature. This requires extra scrutiny and thus needs to be diligently tested according to these 5 steps:

  • Poor old unit tests - to test general 'obvious' cases and edge cases as well as regressions
  • Property based testing on valid json - tests if valid but random generated json parses the same in simd-json and in serde-json (floats here are excluded since slighty different parsing algorihtms lead to slighty different results)
  • Property based testing on random 'human readable' data - make sure that randomly generated sequences of printable characters don't panic or crash the parser (they might and often error so - they are not valid json!)
  • Property based testing on random byte sequences - make sure that no random set of bytes will crash the parser
  • Fuzzing (using afl) - fuzz based on upstream simd pass/fail cases

This certainly doesn't ensure complete safety but it does go a long way.

Other interesting things

There are also bindings for simdjson available here


simdjson-rs itself is licensed under either of

However it ports a lot of code from simdjson so their work and copyright on that should be respected along side.

The serde integration is based on their example and serde-json so again, their copyright should as well be respected.


~27K SLoC