Hacker News Reader

It is confusing. They meant the benchmark used a 500ms "workload" (sleep) per request handler -- x isn't really an axis.

I dont want to interrupt you guys, just writing this info. I will not try to do criticism on other comments, you're free to tell everything about the project, but I think you should read the README first and navigate to the benchmark suite (which is not mine) https://github.com/smallnest/go-web-framework-benchmark . He explains why this is the most realistic benchmark suite so far. Thanks you for using (or talking about) Iris!

The graph at their GitHub page might be a bit easier to understand: https://github.com/kataras/iris

There's a bit more detailed bench here: https://github.com/smallnest/go-web-framework-benchmark