Performance, generally much faster. But they are approaching the problem from different perspectives. Nlohmann focuses on an owning dom like value, JSON Link focuses on parsing into user data structures, but has the ability to mix with a non owning pull parser when if makes sense.
I haven’t added it to that bench but it does it in about 4ms for the 3 files
I haven’t added it to that bench but it does it in about 4ms for the 3 files