Benchmarks

tsz has focused on single-file performance so far. Work is underway to make it fast for full projects too.

Summary

show runner info

Generated 2026-06-21T02:19:14Z · sha 01bebe97d9bd · tsz release-pgo, 17 PGO training inputs, profile 8faa8aec6d05 · linux/x64 · 32 CPUs Intel(R) Xeon(R) CPU @ 2.20GHz · 31.4 GiB RAM · CI runner · Cloud Build e2-highcpu-32

Sum across 78 successful micro benchmark cases.

tsz
23s
tsgo
40s
tsz 1.74x faster
tsz (Rust compiler) tsgo (Go compiler)

Full Project Type Checking

Full-project rows use real repositories and generated app fixtures.

Applications

Pinned real-world applications checked with their own project configuration.

0 lines · 0 KB

tsz
11ms
tsgo
274ms
tsz 24.9x faster
View details

External libraries

20,151 lines · 742 KB

tsz
58ms
tsgo
173ms
tsz 3.0x faster
View details

16,758 lines · 467 KB

tsz
52ms
tsgo
110ms
tsz 2.1x faster
View details

1,012 lines · 27 KB

tsz
75ms
tsgo
125ms
tsz 1.7x faster
View details

3,903 lines · 87 KB

tsz
102ms
tsgo
127ms
tsz 1.2x faster
View details

814 lines · 22 KB

tsz
298ms
tsgo
138ms
tsgo 2.2x faster
View details

8,044 lines · 216 KB

tsz
895ms
tsgo
342ms
tsgo 2.6x faster
View details

Large repositories

222,594 lines · 6,946 KB

tsz
223ms
tsgo
462ms
tsz 2.1x faster
View details

Generated apps

Programmatically created app projects with framework defaults and common TypeScript dependencies.

86 lines · 2 KB

tsz
169ms
tsgo
129ms
tsgo 1.3x faster
View details

282 lines · 7 KB

tsz
176ms
tsgo
133ms
tsgo 1.3x faster
View details

Compile canaries and incomplete project timings

Rows that are tracked for compile readiness but are not part of the timed vs-tsgo chart yet.

Micro Benchmarks

Focused cases for specific compiler paths: single-file library checks, generated type workloads, and solver stress tests.