AI Feedback System Test Results with Jev
Jev was tested in an AI customer feedback system using 1,040 GitHub issues, proving to be 98% cheaper and 84% faster than Claude Sonnet 4.6 while maintaining higher accuracy.
view details →JEV use cases
94 published use cases that name JEV.
Tax Document Classifier Achieves 100% Accuracy
Developed a tax document classifier that processes thousands of documents with 100% accuracy at a cost of $0.001 per page, significantly cheaper and faster than previous methods.
view details →Jev: Decision-Making Model for Structured Outputs
Jev is a model that processes states and typed questions to return structured JSON with probability distributions. It outperforms previous setups in speed and cost, making it ideal for routing and decision-making tasks.
view details →Chrome Extension for Custom X Algorithm
This Chrome extension allows users to create their own X algorithm to filter out engagement bait, empty hooks, and ads. It utilizes Jev by @typesafeai to check posts quickly.
view details →Benchmarking Jev for Automated Code Reviews
The Jev model was benchmarked against an ensemble of models for code reviews, achieving zero false positives, a review speed increase of ~50x, and a cost reduction of ~100x. It demonstrated a 75% bug recall rate, highlighting its efficiency compared to traditional multi-turn agent workflows.
view details →Feed Optimizer for Relevant Posts
This feed optimizer uses a single call to Jev to classify posts into categories like 'ragebait' and 'useful.' It scored 150 posts in under 10 seconds, achieving a 63% retention rate on unseen bookmarks.
view details →Good Night Message
A simple good night message shared with a moon emoji.
view details →Chrome Extension for Real-Time Malicious Address Detection
Jevscan is a Chrome extension that detects malicious addresses and transactions on Etherscan in real-time, utilizing Jev from Typesafe AI.
view details →TypeSafe Plugin for Hermes Agent
This plugin transforms model output into typed judgments: Choice, Score, or Noul, ensuring validation before host code execution.
view details →Simulation of Decision-Making in a Virtual Town
Experience a virtual town where a single entity makes over 20,000 decisions in minutes, managing everything from work to health for 120 residents. Encounter challenges like sickness and disasters as the town evolves.
view details →JevRouter for Hono: Meaning-Based HTTP Routing
JevRouter routes HTTP requests based on meaning rather than method and path. It is available for installation via npm with the package name 'hono-jev-router'.
view details →AI Agent Integration with Jev by Typesafeai
An experiment integrating Jev into a SaaS AI Agent shows remarkable speed, responding faster than user input. It operates at approximately 100 times lower cost than traditional LLMs and efficiently directs requests needing a full LLM.
view details →Jev Integration for Orus Agents
Jev, a model from TypeSafe AI, is now integrated into Orus agents to review proposed trades. It can approve, refuse, or request further observation based on user-defined conditions.
view details →Jev by typesafeai Plays Pokémon FireRed
Jev successfully played Pokémon FireRed and defeated the Elite 4 using a party of level 40 Pokémon. The entire process cost less than $0.03.
view details →AI Reply Detection Tool
This tool flags AI-generated replies on X by analyzing common writing signals. It provides instant probability assessments for each reply through a Chrome extension.
view details →Chrome Extension for Skipping YouTube Sponsor Segments
This Chrome extension listens to YouTube audio and detects sponsor segments to skip them in real-time. It is a prototype project that costs approximately $0.005 per video and is open-source.
view details →Argument Validator Performance Comparison
The argument validator in PatchOpsAi, built with Luna, shows performance metrics of 1.1–2.8 seconds, while @typesafeai's Jev took 195–251 ms on the validator and 221–901 ms on the scanner.
view details →Space Race Game Development with TypeSafe
A space race game was developed for TypeSafe/Jev, achieving 31,382 points with 180 stars and 18 kills in the final 180 seconds. The replay video showcases the decision-making process and verified outcomes without new API calls.
view details →Fine-tuning GLiNER 2.5 Outperforms Jev
Fine-tuning the open-weight GLiNER 2.5 model on a local CPU improved accuracy by ~18pp, surpassing Jev by ~9pp. The process took about 51 minutes and resulted in an 8-10x speed boost for specific tasks.
view details →Chrome Extension for X Timeline Post Labeling
This Chrome extension uses TypeSafe’s Jev model to label posts in your X timeline, identifying ragebait, hype trains, hidden ads, spam, payout farming, AI content, and genuine posts with probability breakdowns.
view details →Trading Bot for AAPL, TSLA, NVDA
This trading bot monitors AAPL, TSLA, and NVDA in USDG, providing buy, sell, or hold recommendations through a single call. Built as a fun project using 3Route, Jev (TypeSafe), x402, and Robinhood Chain.
view details →Improved Model Routing and Action Review with Jev
The integration of Jev from @typesafe_ai replaced three steps in Prio, achieving 100% accuracy in model routing and significantly reducing action review time from 5 seconds to 0.25 seconds for clear cases.
view details →Fast Computer Use Solution with Jev + Cua Driver
The Jev + Cua Driver, known as jev-use, is now available in development preview for macOS, Windows, and Linux.
view details →Evaluation of Jev Model Against LLMs
Early experiments compare the Jev model from Typesafe AI with LLMs like GPT-OSS and GLM. Results indicate Jev may be suitable for production workloads, offering competitive pricing and performance.
view details →