The value prop isn't serde_json, it's automatically generated serializers and deserializers for structured data without needing an extra codegen step like with protobufs/capnproto, plus all that machinery decoupled from the actual data format you're reading.

It essentially generates a massive amount of code that you need to write anyway, at the cost of code size and compile time. And a lot of people are happy to make that trade off.

I wouldn't call that a "bloated monster" because of that. Also, none of those options are alternatives to serde_json, unless you restrict yourself to serde_json::Value - which no one does in practice.