{"slug": "let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe", "title": "Let an AI agent drive your Mac for a fiftieth of a cent per step with TypeSafe Computer Use", "summary": "A developer released TypeSafe Computer Use (jev), an open-source macOS computer-use agent that reads the screen deterministically and routes each step through a small classifier, calling a frontier model only when free text or screen reading is needed. Per the project's README, jev costs about $0.0002 per decision versus $0.032 for Claude Opus 5 on a bare screenshot, and roughly $0.003 for a 12-step task versus $0.40 to $0.90, with model latency of 0.13 to 0.38 seconds against 5.2 seconds. The README acknowledges the trade-off: reasoning a frontier model performs for free must be rebuilt as deterministic state, such as date parsing written in code.", "body_md": "Computer-use agents built on frontier models work, but every step ships a screenshot to a big model and waits several seconds for a plan. A twelve-step task can cost close to a dollar. Most of those steps do not need a plan at all: they need one choice from a short list of buttons.\n\n[awlevin/typesafe-computer-use](https://github.com/awlevin/typesafe-computer-use), called jev for short, is built on that observation. Per the README it drives a Mac toward a goal you type in plain English for about a fiftieth of a cent per step. It reads the screen deterministically, asks a small classifier which action comes next, and only calls a writing model when a text field genuinely needs free text or the classifier has stopped and the screen needs reading.\n\nThis post covers what it is good for, the install, and a safe first run.\n\nThe README measures one decision on the same screenshot and goal against Claude Opus 5 given a bare screenshot:\n\n|  | jev | frontier model, bare screenshot | \n|---|---|---|\n| cost per decision | $0.0002 | $0.032 | \n| cost per 12-step task | $0.003 | $0.40 to $0.90 | \n| model latency | 0.13 to 0.38 s | 5.2 s | \n| full step with capture and OCR | about 1.5 s | about 5.5 s | \n\nIt is honest about the catch: the big model read event dates off the pixels and compared them unaided, while the classifier needed date parsing written in code. Every piece of reasoning a frontier model does for free has to be rebuilt as deterministic state.\n\nYou need macOS 14 or newer, Python 3.12 or newer and [uv](https://docs.astral.sh/uv/). Windows 10 and 11 are experimental.\n\n```\ngit clone https://github.com/awlevin/typesafe-computer-use\ncd typesafe-computer-use\nuv sync\ncp .env.example .env     # fill in the keys\n```\n\nThe `.env` variables:\n\n`TYPESAFE_API_KEY` (required): used for every decision.`ANTHROPIC_API_KEY` (optional): used for typing free text, writer-proposed URLs and the final answer. The README documents pointing the writer at another model instead, including LM Studio, Ollama or any OpenAI-compatible server.`CLICKER_EMAIL` (optional): enables the `type_email` action.`CLICKER_BROWSER` (optional): defaults to Google Chrome.\nThen grant your terminal **Screen Recording** and **Accessibility** in System Settings, Privacy & Security. Without the first, captures are just wallpaper. Without the second, clicks are silently dropped and `--act` refuses to start.\n\nIt drives your real mouse and keyboard, so start without `--act`:\n\n```\nuv run clicker \"open the Playground\"                 # dry run: one step, prints what it would do\nuv run clicker-inspect \"any goal\"                    # capture and open the annotated screen + payload\n```\n\nClear the terminal before you run it. The terminal is on screen, so its text becomes OCR input.\n\n```\nuv run clicker \"open the Playground\" --act           # up to 100 steps\nuv run clicker \"log in\" --act --steps 20 --delay 3   # longer and slower\nuv run clicker \"log in\" --act --handoffs 0           # the classifier alone\n```\n\nTo stop a live run, press Ctrl-C or slam the mouse into the top-left corner of the screen. The loop also stops on `done`, low confidence, a stall or the step limit. Every run writes a `runs/<timestamp>/` folder, so a stalled run can be replayed offline, and the README says a run folder attached to a bug report is the most useful thing you can send.\n\nThe projects collected on the [TypeSafe Computer Use page](https://dedupedai.com/use-cases/awlevin-typesafe-computer-use/?utm_source=devto&utm_medium=article&utm_campaign=howto_typesafe_cu) show it as a base layer:\n\nIf your task lives entirely in a browser, Browser Use's [Jev Ultrafast](https://dedupedai.com/use-cases/browser-use-jev-ultrafast/?utm_source=devto&utm_medium=article&utm_campaign=howto_typesafe_cu) applies the same classifier idea to a numbered table of page elements, and jev itself has a browser backend that reads the DOM over Chrome DevTools with no OCR and no screen permission.\n\n`--steps` value and, for anything risky, the documented Linux sandbox in a container.\nSource repo: [awlevin/typesafe-computer-use](https://github.com/awlevin/typesafe-computer-use). License: MIT.\n\nWant the setup commands for 50 AI repos people actually ship with? Get the free AI Repos That Actually Ship pack: [https://dedupedai.com/ship-pack/?utm_source=devto&utm_medium=article&utm_campaign=shippack](https://dedupedai.com/ship-pack/?utm_source=devto&utm_medium=article&utm_campaign=shippack)", "url": "https://wpnews.pro/news/let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe", "canonical_source": "https://dev.to/dedupedai/let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe-computer-use-1ag", "published_at": "2026-11-04 04:10:51+00:00", "updated_at": "2026-11-04 04:22:09.673294+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "developer-tools", "ai-products"], "entities": ["TypeSafe Computer Use", "jev", "awlevin", "Claude Opus 5", "Anthropic", "Browser Use", "Jev Ultrafast", "LM Studio"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe", "markdown": "https://wpnews.pro/news/let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe.md", "text": "https://wpnews.pro/news/let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe.txt", "jsonld": "https://wpnews.pro/news/let-an-ai-agent-drive-your-mac-for-a-fiftieth-of-a-cent-per-step-with-typesafe.jsonld"}}