89 curated builds·evidence labeled
COMMUNITY & PRODUCTION DIRECTORY

Real Jev Agent Implementations

Live demonstrations, open-source repositories, and sourced benchmarks where developers put Jev into production loops. Filter by architecture pattern or explore video breakdowns below.

Showing 45 sourced implementations (video)
Clear filters ×
JO
Joey Kudish
Jev Browser
DEMO

Jev Browser

A Jev-driven browser agent exposed as an MCP server, CLI, or library.

Latency: ~4s end-to-end for that reported exampleCost: $0.0016 for one reported Wikipedia navigation example
Browser AgentsJevPlaywrightGitHub
TypeSafe AITypeSafe AI@typesafeai
Wikiracing Fast Navigation
DEMO

Wikiracing Fast Navigation

An official benchmark demonstrating Jev selecting optimal Wikipedia links in real time to reach target concepts.

Hop Latency: <50ms / stepBenchmark: WikiracingAction Space: 140+ links
Browser AgentsTypeSafe System One APIChoice
CU
Cua
CUA-S1 Form Automation
DEMO

CUA-S1 Form Automation

An open 706k-parameter System 1 model specialized for mapping user data into web form inputs in 50ms.

Latency: ~50 ms per form fill pass
Browser AgentsCua S1PlaywrightGitHub
Gregor ZunicGregor Zunic@gregpr07
Flight Search with Browser Use
DEMO

Flight Search with Browser Use

A browser agent searching real-world flights at 1× speed with Jev choosing each next action.

Run time: ~7 sCost: ~$0.004
Browser AgentsBrowser UseJev ChoiceGitHub
ClineCline@cline
jev-browser in Cline
DEMO

jev-browser in Cline

A marketplace plugin that connects Jev-driven browser execution directly into the Cline coding agent.

Browser AgentsClineVS Code / Desktop
Andrew LevinAndrew Levin@awlevin
TypeSafe Computer Use
DEMO

TypeSafe Computer Use

Open-source macOS desktop control loop with bounded Jev decisions at every single step.

Cost per step: ~$0.0002Latency: <50ms
Browser AgentsmacOSPythonGitHub
RanjanRanjan@manofsteel3129
AskJev (MCP Agent)
DEMO

AskJev (MCP Agent)

Claude handles conversational intent while Jev directs browser navigation across live websites.

Browser AgentsClaude DesktopMCP ServerGitHub
Sawyer HoodSawyer Hood@sawyerhood
Prompt Box Router
DEMO

Prompt Box Router

Auto-dispatches single prompt inputs to the optimal agent runtime, model tier, target computer, and directory.

Agent RoutingJev ChoiceNode.js / TypeScript
InstructaInstructa@instructa
Switchloom
DEMO

Switchloom

Deterministic model routing for coding agents, featuring an installable skill for Codex.

Correct: 96/100Volume: 1,500 testsSpeed: 100 in 1.42s
Agent RoutingCodex SkillJev ChoiceGitHub
John YeoJohn Yeo@johnyeo_
Slack Agent Skill Routing
DEMO

Slack Agent Skill Routing

Pre-selects skills, tools, and parameters before an enterprise Slack agent runs, doubling speed.

Speed-up:
Agent RoutingSlack BoltJev Choice
jackbarunzjackbarunz@jackbarunz
Jev Tool Router
DEMO

Jev Tool Router

Selects the exact MCP tool for Codex to prevent tool overload and context degradation.

GitHub Stars: 4Decision Latency: ~40ms
Tool SelectionCodexMCPGitHub
Niaz MorshedNiaz Morshed@niazmorshed_
Jev Review (MCP Plugin)
DEMO

Jev Review (MCP Plugin)

An MCP plugin that scores and critiques coding agent output during active iteration loops.

Latency: 129 rows in ~1 s
Agent EvaluationMCP PluginCursor / Claude Code
Raihan KhanRaihan Khan@raihankhan_rk
DiffJury
DEMO

DiffJury

Evaluates public PR links to provide instant safe-to-merge or review-required verdicts.

Agent EvaluationGitHub APIJev Choice
ClownwareClownware@clownware
Bouncer
DEMO

Bouncer

A judgment layer checking every Claude Code tool call against customizable organizational policies.

Cost: ~$0.04/dayPolicy Check: 100%
GuardrailsClaude CodeJev ChoiceGitHub
Idan LevinIdan Levin@0xidanlevin
Jev on WebMCP Benchmark
DEMO

Jev on WebMCP Benchmark

Solved 49 of 49 benchmark tasks at 112× lower model cost by combining Jev tool selection with fast LLMs.

Tasks solved: 49/49Model cost: ~112× lower
Tool SelectionWebMCPMercury 2.5
Kyle JeongKyle Jeong@kylejeong
Stagehand on a remote browser
DEMO

Stagehand on a remote browser

we built blazing fast computer/browser use with Jev + @Stagehanddev.

Cost per task: ~$0.001
Browser AgentsJev APITypeScript / PythonGitHub
Jarrod WattsJarrod Watts@jarrodwatts
jev-trader
DEMO

jev-trader

I built a trading bot with Jev!

Decision window: 300 ms blocks
Agent RoutingJev APITypeScript / PythonGitHub
vogelvogel@ryanvogel
Triage across 1,500 emails
DEMO

Triage across 1,500 emails

this model is actually insane at email classification

Volume: ~1,500 emails
Agent RoutingJev APITypeScript / Python
HassanHassan@nutlope
1kpapers
DEMO

1kpapers

I used Jev to classify 1,018 AI research papers.

Documents: 1,018 papersTotal cost: $0.08Median latency: 256 ms
Browser AgentsJev APITypeScript / Python
Andy GaoAndy Gao@instantricecook
Voice-controlled computer use on a Mac
DEMO

Voice-controlled computer use on a Mac

I built a voice controlled computer-use for my mac using @typesafeai's Jev and it's INSANE how fast it is!

Browser AgentsJev APITypeScript / Python
Jason LuJason Lu@jasonlu_ai
jev-e2e
DEMO

jev-e2e

JEV is changes the world of E2E testing!

Jev: 47 s / $0.0067Sonnet 5: 79 s / $0.4062
Agent EvaluationJev APITypeScript / PythonGitHub
Rob HallamRob Hallam@robj3d3
Post scoring with SuperX
DEMO

Post scoring with SuperX

Jev + SuperX = virality solved ✅

Questions per draft: 61Latency: ~1 sCost: $0.0004
Agent RoutingJev APITypeScript / Python
Matthew BermanMatthew Berman@TheMattBerman
724 competitor ads, broken down
DEMO

724 competitor ads, broken down

jev is INSANE.

Ads: 724 from 37 brandsRun time: ~40 sCost: ~$0.09
Agent EvaluationJev APITypeScript / Python
RomànRomàn@romanbuildsaas
700 leads scored in 40 seconds
DEMO

700 leads scored in 40 seconds

JEV is INSANE.

Leads: 700Time: 40 sCost: $0.09
Agent RoutingJev APITypeScript / Python
Faadil ShaikFaadil Shaik@faadilhshaik
Jev plays Super Mario Bros.
DEMO

Jev plays Super Mario Bros.

got @typesafeai’s new model Jev to play Super Mario Bros.

Tool SelectionJev APITypeScript / Python
Jon KraayenbrinkJon Kraayenbrink@kraayenJon
AI slop detector
DEMO

AI slop detector

jev is INSANE.

Check time: 243 msTells checked: 35Cost: $0.00015
GuardrailsJev APITypeScript / Python
Marcel Pociot 🧪Marcel Pociot 🧪@marcelpociot
Hide posts on X in plain language
DEMO

Hide posts on X in plain language

I built a browser extension with Jev @typesafeai that can hide/collapse posts on

GuardrailsJev APITypeScript / Python
Ian NuttallIan Nuttall@iannuttall
3,282 posts, eight questions each
DEMO

3,282 posts, eight questions each

I gave Jev 3,282 of my X posts across 100M views and asked it to find what actua

Posts: 3,282Cost: $0.1282Run time: 8 m 34 s
Agent EvaluationJev APITypeScript / Python
Riley BrownRiley Brown@rileybrown
500 emails for 3.5 cents
DEMO

500 emails for 3.5 cents

Yeah Jev by @typesafeai is very cool. It classified 500 emails in seconds. And i

Emails: 500Cost: 3.5 cents
Agent RoutingJev APITypeScript / Python
AbolAbol@abolbuild
$10,000 in Jev’s hands
DEMO

$10,000 in Jev’s hands

I gave Jev $10,000 and let it trade

Bankroll: $10,000
Jev APITypeScript / Python
leoleo@leojrr
The X algorithm, rebuilt with Jev
DEMO

The X algorithm, rebuilt with Jev

rebuilt the X algorithm with Jev

Agent EvaluationJev APITypeScript / Python
Tony DinhTony Dinh@tdinh_me
YouTube sponsor skipper
DEMO

YouTube sponsor skipper

Just trying out Jev, I made a Chrome extension that:

Cost per video: ~$0.005
Browser AgentsJev APITypeScript / PythonGitHub
Ian NuttallIan Nuttall@iannuttall
A filter for reply-guy comments
DEMO

A filter for reply-guy comments

Pro tip: You can use Jev to remove annoying reply guy comments that X seems to a

GuardrailsJev APITypeScript / Python
Riley BrownRiley Brown@rileybrown
Live viral post analyzer
DEMO

Live viral post analyzer

Just created this with Jev by @typesafeai. A live viral post analyzer. As soon a

Agent EvaluationJev APITypeScript / Python
Diogo AlmeidaDiogo Almeida@CompleteSkeptic
Jev plays Doom
DEMO

Jev plays Doom

We love how this doomo doomonstrates real-time intelligence and what can be doon

Rate: ~10 queries/sCost: ~$7/hour
Tool SelectionJev APITypeScript / Python
CJ (Syntax.fm)CJ (Syntax.fm)@syntaxfm
wtf is jev? Demos: Browser Use, Classifier, Router
DEMO

wtf is jev? Demos: Browser Use, Classifier, Router

Demonstrates browser automation, model routing, and tokenless agent loops.

Format: YouTube 4KDemos: 5 live setups
Browser Agents
Greg IsenbergGreg Isenberg@GregIsenberg
Jev is HERE. How to use it
DEMO

Jev is HERE. How to use it

High-volume business workflow automation via Vercel AI Gateway.

Podcast: Startup IdeasTopic: AI Business & Gateway
Agent Routing
Ryan VogelRyan Vogel@vogeldev
Meet Jev: tested on 1,000 real emails
DEMO

Meet Jev: tested on 1,000 real emails

Selects category, urgency level, and auto-reply trigger in single pass.

Volume: 1,000 emailsTotal Cost: <$0.01
Tool Selection
AICodeKingAICodeKing@AICodeKing
Jev and Browser Use, fully tested
DEMO

Jev and Browser Use, fully tested

Validates incoming actions against injection patterns before navigation.

Domain: Browser AutomationSafety: Injection tested
Browser Agents
Ray AmjadRay Amjad@RAmjad
Jev + Claude Code: the cheapest agentic loop yet
DEMO

Jev + Claude Code: the cheapest agentic loop yet

Decides whether to re-read files, run tests, or commit changes.

Loop Cost: 80% cheaperStack: Claude Code + Jev
Agent Evaluation
Prompt WarriorPrompt Warrior@promptwarrior
Jev full tutorial: 3 prototypes built live
DEMO

Jev full tutorial: 3 prototypes built live

Scores relevance and routes between local voice input and browser execution.

Prototypes: 3 live buildsLevel: Step-by-step
Browser Agents
@latentspacepod@latentspacepod@latentspacepod
Why I Couldn't Build Jev at OpenAI: Diogo Almeida
DEMO

Why I Couldn't Build Jev at OpenAI: Diogo Almeida

TypeSafe Co-founder & CEO Diogo Almeida explains why post-training on generative LLMs fails on agent loops and how RLCD trains System 1.

Speaker: Diogo AlmeidaDuration: 58 minsTopic: RLCD & System 1
Agent RoutingTypeSafe System OneRLCD
PatCodesPatCodes@patvscode
Open-Source Jev? Install Laya Locally + 3 Useful Demos
DEMO

Open-Source Jev? Install Laya Locally + 3 Useful Demos

Step-by-step tutorial installing the open-source 420M parameter Laya model locally on CPU via ONNX with 3 runnable demos.

Runtime: ONNX / CPUModel Size: 420M paramsLatency: 8~12ms
MCP / SkillsPythonONNX Runtime
Marcin Zajkowski (What’s NEXT)Marcin Zajkowski (What’s NEXT)@zajkowskimarcin
What's NEXT #9: Smaller AI, Local Models & Jev
DEMO

What's NEXT #9: Smaller AI, Local Models & Jev

A deep discussion on why enterprise architectures are slimming down from heavyweight LLMs to fast specialized decision nodes.

Topic: Enterprise AIDuration: 22 minsFocus: Multi-Agent Bus
Agent RoutingTypeSafe System OneSmall Models
Patrick DangPatrick Dang@PatrickDang
Live Sales Role Play & Dynamic Objection Triage
DEMO

Live Sales Role Play & Dynamic Objection Triage

A sales training and live role play agent that categorizes customer objections in real time to trigger battle cards.

Domain: Sales CoachingLatency: <50msDuration: 9 mins
Agent EvaluationTypeSafe System OneLive Audio Pipeline