Web framework comparison: any two entries, any language
cinatra is ahead of ntex on 10 of the 19 tests they both run, and behind on 9. On the HTTP/1.1 composite ntex scores 6613 and cinatra 6349.
Pick any two of the 212 entries, in any language - the picker is grouped by language, and the two sides do not have to match. Requests per second is the best of three runs and the delta is the first column against the second; every row is the same profile at the same connection count, on the same machine. The fixed-rate profiles offer every entry the same request rate, so they are compared on their score out of 100 instead, which is CPU cost and latency at that rate. What each profile measures.
cinatra is ahead of ntex on 10 of the 19 tests they both run, and behind on 9. On the HTTP/1.1 composite ntex scores 6613 and cinatra 6349. Both run on the same machine, in the same round, against the same profiles.
Is ntex faster than actix?
ntex is ahead of actix on 21 of the 31 tests they both run, and behind on 10. On the HTTP/1.1 composite ntex scores 6613 and actix 6322. Both run on the same machine, in the same round, against the same profiles.
Is ntex faster than fiber-tuned?
ntex is ahead of fiber-tuned on 12 of the 13 tests they both run, and behind on 1. On the HTTP/1.1 composite ntex scores 6613 and fiber-tuned 5859. Both run on the same machine, in the same round, against the same profiles.
Is cinatra faster than actix?
cinatra is ahead of actix on 13 of the 19 tests they both run, and behind on 6. On the HTTP/1.1 composite cinatra scores 6349 and actix 6322. Both run on the same machine, in the same round, against the same profiles.
Which framework is fastest overall?
Among the web frameworks, dusty-tuned holds the highest HTTP/1.1 composite, at 7638; among the HTTP engines, edixoi-kt holds the highest HTTP/1.1 composite, at 6729; among the experimental frameworks, viltrum holds the highest HTTP/1.1 composite, at 7858; among the reverse proxies and static-file servers, agensio holds the highest HTTP/1.1 composite, at 4507. Each league is measured against its own leader, and the leagues do not all count the same profiles, so those numbers are not on one scale and the highest of them is not the fastest entry on the page: to set an entry of one league against an entry of another, pick the two in the table above. The composite is throughput over the whole suite; if a service is one shape of workload, compare that profile on its own in the table above, because the order changes.
Can I compare frameworks written in different languages?
Yes - that is what this page is for. Every entry runs the same profiles on the same machine, so an F# entry and a C# one are measured the same way and the numbers line up. The per-language pages under Comparisons narrow the same table to one language when that is the choice you are making.