I'm not too familiar with this project but based on the linked test which is most realistic for this type of project https://klen.github.io/py-frameworks-bench/#remote there doesn't seem to be much difference, in fact flask is better than some other frameworks. Is flask really the bottleneck for this project?
In my personal experience, unless you're going with an async type of framework the performance gains from switching from flask to falcon are negligible at best.
The network latency of the foreign search requests is much bigger than the performance penalty of flask. While we can't prove that flask significantly degrades our performance, i have to agree with @Nikola-K.
I forgot the question mark at the end of the subject and the first sentence :-)
Adding to the discussion that there is even a drop in replacement for Flask called Quart which is written with python3's asyncio. Considering python2's end of life, shouldn't we migrate searx to python3 making use of asyncio, uvloop, etc.?
Most helpful comment
The network latency of the foreign search requests is much bigger than the performance penalty of flask. While we can't prove that flask significantly degrades our performance, i have to agree with @Nikola-K.