Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

jev-playground (Ahmar004)

Interactive teaching site where Jev races frontier LLMs in games, an arena, a sandbox and quizzes, with a keyless replay mode.

Appproduction
View on GitHub ↗Live demo ↗

Overview

Jev's Playground is an interactive website that teaches where System One models like Jev work well, where they break, and where a frontier LLM or plain code is the better tool. You learn by playing: levels, 14 games where Jev races an LLM, a side-by-side Arena, a hands-on Sandbox, quizzes and a leaderboard.

Beginner mode replays real recorded runs, so no API key is needed; Developer mode runs everything live with your own keys. The README documents how keys are handled, the tech stack, numbers gathered so far, trade-offs and how AI was used to build it. Version 2.0 is live and contributions are welcome.

What the decision model decides

TypeQuestionAnswers
OtherGame and sandbox tasks run through Jev and an LLM side by side(inferred)—

How it works

  1. Open the site and choose a level, game, Arena or Sandbox
  2. Beginner mode replays recorded runs
  3. Developer mode calls Jev and LLMs live with your keys
  4. Compare results and take quizzes

Strengths

  • No setup needed in Beginner mode
  • Shows weaknesses as well as strengths
  • Hands-on comparison with frontier LLMs

Caveats

  • Educational, not a rigorous benchmark
  • Live mode needs your own keys

Getting started

Visit https://letsplaywithjev.vercel.app

At a glance

Type
App
Status
production
Language
TypeScript
Runtime
Web app
Has code
yes

Use cases and integrations

How curated lists describe it

  • A fun and interactive playground to play with Jev and make Jev and frontier LLMs do the same tasks, to learn the use cases of Jev in a fun and interactive way! :).

    hellogumbo/awesome-jev

Frequently asked questions

Do I need an API key?

No. Beginner mode replays recorded runs.

What will I learn?

Where typed decision models beat LLMs or code, and where they fail.

Can I run things live?

Yes, in Developer mode with your own keys.

Related projects

  • warped-sys1-lab

    Local SvelteKit workbench and CLI for organizing, editing, running and comparing Jev experiments defined as JSON.

    CLIbetaTypeScript
  • OpenJev (GPT-AGI)

    Jev-compatible open System One engine with a Claude Code-style terminal REPL showing live probabilities, plus an experiment harness.

    ProjectprototypePython
  • jev-behavior-study

    Independent field guide to Jev 1.13.0 behaviour across 11,621 requests: framing sensitivity, Snake and 3D city driving experiments.

    ProjectresearchPython, TypeScript
  • jev-report

    Independent 52-page Chinese report on Jev that checks vendor claims and reproduces 50 Chinese tests, with traceable data.

    ProjectresearchPython
  • typesafe-showcase★ 1

    Next.js site with six live Jev demos, each compared side by side with Gemini or Claude on latency, cost and agreement.

    DemobetaTypeScript
  • jev-playground

    Local single-page playground for Jev with templates from 37 benchmark datasets, showing gold answers next to typed results.

    AppprototypePython