Skip to content

docs: T3 Code benchmark against bun check and tsc 7 #6

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
t3dotgg merged 1 commit into from
Oct 7, 2026
Merged

docs: T3 Code benchmark against bun check and tsc 7 #6

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 33 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,6 +95,39 @@ that revision:
- **Real projects.** On 120 open-source repos, the command-line output differs from Go's only in
the problems below and where Go's own output changes from run to run.

## Benchmark: T3 Code

Full type check of [T3 Code](https://github.com/pingdotgg/t3code), compared with `tsc` 7 and
the new `bun check` in Bun. T3 Code uses Effect, so there are two cases: without the Effect
diagnostics and with them. The multiplier compares with `tsc` 7 in the same case. Lower is faster.

**Without Effect diagnostics**

| Checker | Time | vs `tsc` 7 | |
| ----------- | -----: | -------------- | ------------------------------------- |
| `bun check` | 4.21s | 4.09脳 faster | `鈻堚枅鈻堚枅` |
| `tsc-rs` | 7.70s | 2.24脳 faster | `鈻堚枅鈻堚枅鈻堚枅鈻堚枅` |
| `tsc` 7 | 17.22s | 1.00脳 baseline | `鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻坄 |

**With Effect diagnostics**

| Checker | Time | vs `tsc` 7 + Effect | |
| ----------------------------------- | -----: | ------------------- | --------------------------------------- |
| `tsc-rs` (Effect built in) | 11.35s | 1.94脳 faster | `鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻坄 |
| `tsc` 7 + `@effect/tsgo` | 22.07s | 1.00脳 baseline | `鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅` |
| `bun check` + separate Effect pass | 37.17s | 1.68脳 slower | `鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻堚枅鈻坄 |

`bun check` is the fastest when you do not need the Effect diagnostics. It does not have them, so
an Effect project needs a second pass. `tsc-rs` gets them from its one check.

How it was measured: the sum of per-project medians (3 runs each) for the five T3 Code projects
`apps/server`, `apps/web`, `apps/mobile`, `packages/client-runtime` and `packages/shared`.
Apple M5 Max, warm filesystem cache, compiler caches cleared before each run. This is not a timed
full-workspace or parallel CI run. Versions: `tsc-rs` 0.1.0, TypeScript 7.0.2, Bun canary
`bd599f5af`, `@effect/tsgo` 0.46.1. The results come from separate rounds. The `tsc-rs` switch in
T3 Code is [pingdotgg/t3code#16704](https://github.com/pingdotgg/t3code/pull/16704), with a
per-project table.

## Known problems

- In some monorepos, the source files of a workspace package are reachable both through
Expand Down