Every project here really calls Jev in its source · refreshed daily
questions: { intent: boolean, target: boolean, site: boolean, // +5 }
Control a real browser by voice. Jev (TypeSafe System One) decides intent + target in ~300 ms per spoken word; Playwright acts — often before you finish the sentence.
questions: { color: choice, }
Open-source macOS AI computer use and native GUI automation on Apple silicon. Jev + OmniParser CoreML + Apple Vision OCR. Bring your own OpenRouter, Vercel AI Gateway, or TypesafeAI token.
questions: { unambiguous_match: boolean, }
Policy-bounded Jev target selection for resilient Playwright workflows
questions: { irreversible: boolean, goal_done: boolean, }
AskJev — Jev autopilot for any website + guard on irreversible clicks (TypeSafe System One, not Claude)
questions: { partial: boolean, score: boolean, }
A fast browser agent: Jev picks each action from what is on the page, an LLM reads and plans, and every claim in an answer cites a quote from the page.
questions: { operation: choice, }
Jev for Chrome: drives the tab you are looking at with TypeSafe Jev, a sub-second decision model. Community port of browser-use/jev-ultrafast, not affiliated with TypeSafe.
questions: { operation: choice, click_target: choice, goal_reached: boolean, // +1 }
Fast, bounded browser agents powered by Jev and agent-browser — typed actions, research, classification, and safe orchestration.
questions: { intent: boolean, target: boolean, site: boolean, // +5 }
Türkçe ve İngilizce doğal konuşmayla Windows 10/11 bilgisayar kontrolü: OpenAI Realtime, local Whisper, Jev, UI Automation ve Playwright.
questions: { content_type: choice, }
Fast structured Android control loops with TypeSafe Jev and Mobile MCP
questions: { complete: boolean, }
Open-source native computer use for macOS and Windows: TypeSafe Jev, local OCR, and selective planning.
questions: { completion: choice, }
One grounded Jev/Playwright core: typed SDK, persistent CLI, and MCP server with native browser operations and deterministic assertions.
questions: { action: choice, close_intent: boolean, intentional_control: boolean, // +1 }
Jev-based local-first gesture and gaze media-control agent with guarded page-close intent.
questions: { action: choice, }
Voice and text browser control with Jev, Chrome MV3 and a local AI Gateway bridge
questions: { risky: boolean, submit_after_typing: boolean, }
Trying automation on web browser via jev from typesafe
questions: { refund: boolean, }
Complete browser tasks with Ego Lite and actively call Jev for semantic target selection, filtering, ranking, classification and text evidence judgments.
result = post_json("https://api.typesafe.ai/v1/systemone", os.environ["TYPESAFE_API_KEY"], body)i. am. speed.
typesafe: {endpoint:'https://api.typesafe.ai/v1/systemone',keyName:'TYPESAFE_API_KEY',model:'jev-latest',modelPattern:/^jev-[a-z0-9.-]{1,80}$/},5–10x faster browser operations: Jev clicks, Codex thinks and verifies. Built at EZCollegeApp.
import { TypeSafeClient } from "@typesafe-ai/sdk";Browser use using Typesafe's Jev model
var request = URLRequest(url: URL(string: "https://api.typesafe.ai/v1/systemone")!)
Say it, and your Mac does it. A computer-use harness on Jev that reads the screen through Accessibility. Fast, no vision model
export const API_URL = process.env.JEV_API_URL || "https://api.typesafe.ai/v1/systemone";
Browser automation where an LLM plans and Jev (Typesafe System One) decides. Library, CLI and MCP server.
import { createGateway, experimental_evaluate as evaluate, type Experimental_EvaluationQuestion as EvaluationQuestion } from "ai";Experimental browser agent powered by FX, Jev, and Vercel AI Gateway. Bring your own API key to read pages and automate browser tasks.
const result = (await postJson("https://api.typesafe.ai/v1/systemone", key, body)) as {Fast browser agent for ego lite. One TypeSafe request per step; an agent or Jev picks the move.
import { choice, TypeSafeClient } from '@typesafe-ai/sdk';English Português (Brasil)
process.env.TYPESAFE_BASE_URL ?? "https://api.typesafe.ai",
Browser automation with Jev (TypeSafe) as decision model
from typesafe_sdk import Choice, TypeSafeClient
Aside agents decide with TypeSafe Jev (System One: Choice/Score/Noul). Not a Cua binding — Jev is the model, Aside is the browser runtime.
} else if r.URL.Host != "api.typesafe.ai" {macOS computer use driven by Jev (TypeSafe System One) as the decision maker
"typesafe": "https://api.typesafe.ai/v1/systemone",
A typed Python framework for controlling Android over ADB with Jev.
API_HOST = "api.typesafe.ai"
Pure Jev that can "type" and drive towards task completion.
const TYPESAFE_URL = "https://api.typesafe.ai/v1/systemone";
A small Jev-powered bridge to Chrome through the Chrome DevTools Protocol.
transport: 'typesafe', endpoint: 'https://api.typesafe.ai/v1/systemone',
Browser automation CLI for AI agents, powered by the Jev model's millisecond decisions and near-zero inference costs
result = post_json("https://api.typesafe.ai/v1/systemone", os.environ["TYPESAFE_API_KEY"], body)⚡ Ultrafast browser agent with dynamic, indexed action spaces. Zurich → London on Google Flights in 7.1s.
TYPESAFE_ENDPOINT = "https://api.typesafe.ai/v1/systemone"
Browser use for coding agents, 3-5x faster than browser-use. MCP server + CLI; TypeSafe Jev decides every step in ~300 ms.
URL = "https://api.typesafe.ai/v1/systemone"
@typesafeai 's Jev controls the 2 hands and each finger to play the piano in real-time. Jev only "sees" what we see and plays this from the "note waterfall". It uses @browser_use 's jev-ultrafast and some decision scheduling to make this happen in real-time. Sound on 🔈🔉🔊
baseURL: URL = URL(string: "https://api.typesafe.ai/v1/systemone")!,
Control your Mac by voice. Speech → Jev (TypeSafe AI System One model) typed decisions → macOS actions. Menu-bar Swift app.
* Service worker: the only place that holds the API key and talks to api.typesafe.ai.
Chrome extension that reads every cookie banner and popup like a person and clicks the honest button. Judged by TypeSafe's Jev. BYOK, no backend.
const r = await fetch("https://api.typesafe.ai/v1/systemone", { method: "POST", headers: { Authorization: `Bearer ${KEY}`, "Content-Type": "application/json" },Almond-fastloop: Almond's browser computer-use rig (Chrome DevTools + TypeSafe Jev), and the Browser Use Olympics benchmark it is measured on.
if not model or model == "jev-latest":
Ultra-fast browser automation using TypeSafe Jev via Panerelay and OpenRouter.
const val ENDPOINT = "https://api.typesafe.ai/v1/systemone"
Computer use with TypeSafe Jev: control your PC, browser, games and Android phone with a prompt
const BASE_URL = process.env.TYPESAFE_BASE_URL || "https://api.typesafe.ai/v1";
Drive the ego lite browser with Jev (TypeSafe System One): one indexed element table in, one operation + target out, single process. ~2x faster than a per-step LLM loop in our measurements.
jevModel: options.jevModel ?? (process.env.OPENROUTER_JEV_MODEL?.trim() || 'typesafe/jev-1.13'),
Natural-language end-to-end tests for web apps, powered by Jev and Playwright.
import { TypeSafeClient, choice, noul } from '@typesafe-ai/sdk';Write Cucumber tests with just the .feature file. No step definitions — Jev (TypeSafe AI) resolves each Gherkin step and Playwright runs it.
export const TYPESAFE_ENDPOINT = "https://api.typesafe.ai/v1/systemone";
Isolated Playwright browser for pi, driven by Jev typed decisions through the TypeSafe API or the model pi already has configured.