# NexusFIX Benchmark Evidence

Official page: https://stratcraft.ai/nexusfix/benchmark

No reproducible measurement results are published for the NexusFIX and QuickFIX comparison. Earlier latency, throughput, allocation, hardware, and percentage figures were withdrawn because the original run, exact versions, and measurement method could not be verified. The prior illustrative parse figure was not a measured result.

The [reproduction guide](https://github.com/StratCraftsAI/NexusFix/blob/main/BENCHMARK_REPRODUCTION.md) describes how to run a benchmark. Its expected outputs are not raw evidence for a dated comparison.

A future result must identify the source commits and dependency versions, CPU and operating system, build flags, FIX message corpus, warm-up and affinity, sample count, raw samples, and calculation method. A QuickFIX comparison also requires the same workload and environment for both implementations.

Updated: 2026-09-28