Therefore the Benchmark is always executed for FuzzyWuzzy, RapidFuzz and when directly calling the CPP handler without redirection (e.g. `rapidfuzz.fuzz.fuzz_cpp.ratio`).
Afterwards a ratio between the runtime of both results is calculated. The benchmark can be found [here](https://github.com/rhasspy/rapidfuzz/blob/master/python/bench). The results of the benchmarks are visualised below.