HunterAlphaHub
OpenRouter model reference Facts from the public catalogue, dated and labelled
Union Alpha revealed no live codename Verified
2026-09-22

GitHub · Browser agents

awlevin/typesafe-computer-use

Cost per action treated as a first-class number.

The repository's GitHub card, read on 2026-09-23

What it does

Computer use for about $0.0002 a step: OCR the screen, classify the next action with TypeSafe, click. macOS is the supported platform and Windows is marked experimental in the repository's own badge line.

The repository's own description: “Computer use for about $0.0002 a step: OCR the screen, classify the next action with TypeSafe, click. macOS.”

How it works

`typesafe_computer_use/` is the package, `bench/` holds benchmarks, `tests/` plus a GitHub Actions workflow keep it honest, and uv.lock pins the dependency set — the per-step price is stated in the README rather than buried in a results table.

The repository, by the numbers

Stars
779
Licence
MIT
Stack
Python 3.12 (pyproject.toml + uv.lock), macOS first
Last push
2026-09-21

Read from the GitHub API on 2026-09-22. Stars and the last push move daily — quote them with the date, the way we do.

Numbers the author reports

These are the project's own figures, not our measurements — we keep ours on the statistics page and label them as ours. Read on 2026-09-23, with the sentence each figure appeared in:

302–380 msone full step at p50, of which ~280–350 ms is the decision
“End-to-end loop, p50: 302-380 ms per step (2.6-3.5 steps/sec), of which ~280-350 ms is the TypeSafe decision.” — repository README

What we checked

  • the tree: typesafe_computer_use, bench, tests, docs, .github/workflows, uv.lock
  • MIT LICENSE (Aaron Levin)
  • README's badges say Python 3.12+, macOS (Windows experimental) and points at TypeSafe's docs
  • the per-step cost appears in the repository description and the README

What we did not check. We did not run it — it needs macOS and screen access — so the $0.0002 is the author's figure.

Open the repository ↗All builds