Is the Jev part core to Cua, or one optional recipe?

One optional recipe. The README fetched here covers Cua Spaces, Cua Driver, Lume, the SDK and CUA-S1 with no mention of Jev. The integration is a how-to guide plus example adapters, so Cua Driver is fully usable without TypeSafe credentials.

Does Jev click the mouse, or only pick the action?

It only picks. The recipe has Jev choose among bounded action IDs, and Cua Driver performs the actual clicks and form fills. Execution stays inside the driver, which is what the Python and TypeScript jev_adapter samples demonstrate.

Is 0.23.2 the Cua Driver floor for the jev-use recipe, and do I need a live TypeSafe key?

The how-to guide states Cua Driver 0.23.2 or later, plus Chromium and a TypeSafe console key for live Jev calls. Whether the Python or TypeScript jev_adapter can be pointed at a stubbed judge instead of the hosted model is not documented next to the example, so plan on supplying the key.

Can it drive a real desktop, or only browser pages?

Both, and more. Cua Driver inspects and operates native desktop apps and browsers on macOS, Windows and Linux, with background delivery where the app and platform allow it. The Jev recipe is only the browser form-fill walkthrough.

What is in this repository besides the Jev recipe?

It is the Cua monorepo: Cua Spaces for full agent desktops on your Mac and machines you own, Cua Driver for desktop automation, Lume for local macOS and Linux VMs, CUA-S1 decision models, and Cua Bench for evaluating computer-use agents. Jev is one optional recipe within it.

How do signed-in apps move into a Space?

Teleport moves a signed-in app such as Chrome or Slack into a Space and it opens there still signed in. Sessions stay in the Cua Keyvault, encrypted on your Mac, and reach a Space only after you approve.

How is Cua licensed, and what is free?

The Cua Spaces app is source-available under FSL-1.1-MIT, and Spaces is free for individuals with Pro and Teams plans coming soon. The installer also adds the cua CLI and can wire the driver MCP and skill into your coding agents.

CommunityJev Demos & Experiments

Cua

The Cua computer-use platform documents a jev-use recipe where TypeSafe Jev picks bounded action IDs and Cua Driver executes clicks and fills in the browser.

See highlights

Open github.com

Stars: 29.3k
Language: Rust
License: MIT
Published by: Community
Author: trycua
Last verified: Oct 2026

trycua/cua

Organization repository on GitHub

View on GitHub

Scale computer-use 2.0 with open-source drivers, cross-OS fleets, and benchmarks for training, evaluation, and data generation.

Forks: 2.1k
Last pushed: Updated Oct 2026

Repo stats from the GitHub API, cached Oct 2026.Homepage

About this project

Give AI agents computers they can use. Cua Spaces gives your agents full desktops on your Mac and on machines you own. This repository also holds Cua Driver for desktop automation, Lume for local VMs, CUA-S1 decision models and Cua Bench for evaluating computer-use agents. Cua Spaces: Install the app and run your first Space. Cua Driver: Operate Calculator and verify its result. Lume: Create a Tahoe VM and connect over SSH. Cua SDK and CLI: Install cua and create a local sandbox.

Quoted from the project's README. Read the full README.

Tagscommunity browser computer-use agents

Works forJev for Realtime control Jev for Classification

Used inJev for Software & DevTools

Languages: Python TypeScript
Platforms: Browser Playwright
Capabilities: Open sourceRequires API key

Highlights

docs/content/docs/how-to-guides/driver/jev-use.mdx walks a form-fill demo with Jev choosing actions and Driver submitting.
libs/cua-driver/examples/jev-use/ ships Python and TypeScript jev_adapter samples plus validation docs.
Jev only picks the action; Cua Driver still resolves and performs the click or fill.

Watch out

MIT. Large monorepo; jev-use needs Cua Driver 0.23.2+, Chromium, and a TypeSafe console key for live Jev calls.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

Hacker News: Cua asks which computer-use steps need a frontier model and which are fast decisions.

frabonacci95 pointsShow HN: CUA-S1 – A System One Model for Computer UseRead the thread on Hacker News

X: Jev driving real Chrome from Grok Bot

X: Headless Chromium agent

X: Computer use at 155× cheaper than Opus 5

More Jev resources like this

Fast Browser Use

221

An agent skill that reproduces Jev-style single-step logits over indexed browser controls using local Qwen weights for offline browser use.

Jev Demos & Experiments#community#browser#open-models

Community

APUS-AI-Lab/fast-browser-usePython

CodexQA Jev Browser

A Playwright-based CodexQA skill where Jev chooses indexed controls on a page-built index while the browser executes the action.

Jev Demos & Experiments#community#browser#playwright

Community

openqa-cn/jev-browserTypeScript

surf

634

A CLI browser automation tool with optional Jev semantic mode: Surf observes, Jev selects the next action, Surf validates and acts, Jev checks the goal.

Jev Tools & CLIs#community#browser#cli

Community

nicobailon/surf-cliTypeScript

FluidUse

306

Local macOS computer-use automation on Apple silicon using on-device laya typed decisions and CUA-S1-FORMS for filling, via the Accessibility API.

Jev Demos & Experiments#community#browser#local

Community

FluidInference/FluidUseSwift

Talk to Jev

A chatbot with no language model: every word is picked from roughly 4,700 options by Jev Choice questions, with confidence-driven lookahead, beam search across sentence directions and a self-critique loop that rewrites weak sentences. Code for the paper and the live demo at talktojev.com.

Jev Demos & Experiments#community#chatbot#research

Community

xucian/talktojevPython

FlightBench

Fixed-wing landing simulator where a human can compete against Jev: four Jev Choice decisions map to aileron, elevator, rudder and throttle commands in JSBSim, run in real time. The benchmark also compares Jev, classical controllers and Codex models on landing score and API cost.

Jev Demos & Experiments#community#simulation#aviation

Community

AlperKartkaya/FlightBenchPython

From the Jev guides

Original write-ups on what this entry is for.

Browser Agents and Computer Use: What to Copy from the First Breakout Category

Sep 22, 2026

Browser loops are long series of small decisions, and the expensive model does not need to make any of them. The patterns behind the fastest agents, from element tables to OCR plus accessibility trees.

Jev Tools & CLIs#browser#computer-use#agents

Read article

Per-Tick Decisions in Games

Sep 28, 2026

A game tick is a state plus a legal action list. How Clash Royale, Minecraft, and the earlier Mario, Pokémon, StarCraft, and drone loops ask Jev one step at a time.

Jev Demos & Experiments#gaming#agents#control

Read article

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai