Perf regression: json, json

| Scenario | Environment | Date | Old RPS | New RPS | Change | Deviation | StDev | Dependencies |
| -------- | ----------- | ---- | ------- | ------- | ------ | --------- | ----- | ------------ |
| json | Json Platform 8- Trends gold-lin | 06/26/2026 15:57:18 | 2,099,177 | 2,017,454 | -3.89 % (-81,723) :thumbsdown: | -17.94 σ | 4,554.68 | Changes NameVersionDiff Microsoft.AspNetCore.Antiforgery.dll `11.0.0-preview.6.26323.110` -> `11.0.0-preview.7.26325.123` https://github.com/dotnet/dotnet/compare/ff45a39124929c3f7496f16420d2f0f0265252c4...354f8b358deba86fa37d9fca5ad80a87ac4853f3 Microsoft.AspNetCore.App `11.0.0-preview.6.26323.110` -> `11.0.0-preview.7.26325.123` https://github.com/dotnet/dotnet/compare/ff45a39124929c3f7496f16420d2f0f0265252c4...354f8b358deba86fa37d9fca5ad80a87ac4853f3 Microsoft.NETCore.App `11.0.0-preview.6.26323.110` -> `11.0.0-preview.7.26325.123` https://github.com/dotnet/dotnet/compare/ff45a39124929c3f7496f16420d2f0f0265252c4...354f8b358deba86fa37d9fca5ad80a87ac4853f3 |
| json | Json 8- Trends gold-lin | 06/28/2026 13:35:25 | 1,672,098 | 1,578,915 | -5.57 % (-93,183) :thumbsdown: | -8.11 σ | 11,495.08 | Changes NameVersionDiff Microsoft.AspNetCore.Antiforgery.dll `11.0.0-preview.7.26325.123` -> `11.0.0-preview.7.26327.102` https://github.com/dotnet/dotnet/compare/354f8b358deba86fa37d9fca5ad80a87ac4853f3...3e328deefbd3029c2f957050c508b4faeee3d71f Microsoft.AspNetCore.App `11.0.0-preview.7.26325.123` -> `11.0.0-preview.7.26327.102` https://github.com/dotnet/dotnet/compare/354f8b358deba86fa37d9fca5ad80a87ac4853f3...3e328deefbd3029c2f957050c508b4faeee3d71f Microsoft.NETCore.App `11.0.0-preview.7.26325.123` -> `11.0.0-preview.7.26327.102` https://github.com/dotnet/dotnet/compare/354f8b358deba86fa37d9fca5ad80a87ac4853f3...3e328deefbd3029c2f957050c508b4faeee3d71f |

[PowerBI Dashboard](https://aka.ms/aspnet/benchmarks)

@sebastienros

Start with the JSON benchmark results in this issue and the linked PowerBI dashboard, then compare the listed Microsoft.AspNetCore.App and Microsoft.NETCore.App dependency changes between the reported builds. No source file or test is named, so first identify the benchmark entry point and the change associated with the RPS drop. Done means the regression is explained and the benchmark result is validated after the relevant change.

Written by the indexing model from the issue text.