| Clang Parse |
| Clang Serialize |
| GCC Parse |
| GCC Serialize |
| Benchmarks are computed by running "./bench -i:b *.json", and comparing the results of the pull request with the develop branch. The margin of error for each test is usually in the range of 1%. However, it can be less or more. All the charts may be viewed at https://benchmark.cppalliance.org |