We don't maintain a reference benchmark suite compared to other
technologies; unfortunately have also found in general that it is actually
very hard to create benchmarks that give results which are representative
of real world usage.

Broadly, C++Proto is one of the most optimized binary format parsers in the
world though. There's many formats that are faster (including CapnProto and
Flatbuffers, but also many FFI binding style libraries or blind struct
layout writes) which which are objectively faster despite having much less
optimization attention.

Generally it's because in those cases the wire format itself has the
characteristic of near-zero cost parses, which comes at the expense of
schema evolution ability; where schema evolution ability is generally the
top priority for Protobuf because we are intended to be used for long term
evolution in RPC systems where you can't update both sides of the
communication atomically. Basically, cases where you have limited or no
need for schema evolution (because e.g. you are doing FFI within one system
where you update both sides) are the cases where other technologies can
easily beat Protobuf on performance.

On Thu, Feb 19, 2026 at 8:07 AM Pooja Sadevra <[email protected]>
wrote:

> Hi Protobuf team ,
>
> I am wondering if any runtime analysis has been done of protobuf versus
> boost binary serialization ? I saw a really old thread ( from 2009 ) that
> said text archive is much slower than protobuf . I am wondering if any
> updated analysis has been done on this aspect ? Are protocol buffers more
> runtime efficient when serialising data to binary ?
>
> Thanks
> Pooja
>
> --
> You received this message because you are subscribed to the Google Groups
> "Protocol Buffers" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion visit
> https://groups.google.com/d/msgid/protobuf/b75b5a8a-cfbb-4672-9656-feafffb6ee20n%40googlegroups.com
> <https://groups.google.com/d/msgid/protobuf/b75b5a8a-cfbb-4672-9656-feafffb6ee20n%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"Protocol Buffers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/protobuf/CAKRmVH-qZK7S%2Bc%3DPQRKujiK8ajqZLYAoNLkSg_NNEWoLj%2Bi3gQ%40mail.gmail.com.

Reply via email to