Compares Models on Web Tasks
Gosuke Suzuki tested Jev alongside other models across 500 pages and 20,500 decisions. Jev was the fastest of the eight configurations tested.
Gosuke Suzuki|Suzuki Soten@matu79go𝕏
Same job, five models side by side. Jev was the fastest of all 8 configurations I tested — 500 pages, 20,500 decisions. Code, raw numbers and a browser demo: https://t.co/uUEt19uBJE https://t.co/GvczsRHOYD
The project includes code, raw numbers, and a browser demo.
Also filed under Agents & browsers
- Watchdog for AI Coding Agents
Jev Watch is a watchdog for AI coding agents that stops Codex or OpenCode when they loop, stall, or drift off task, then resumes the same session with a correction.
- Coding-Agent Runtime Using Jev Decisions
Distill is a coding-agent runtime that uses Jev decisions for model routing, context retention, and tool selection.
- Jevry Desktop Browser
Jevry is a desktop browser where Jev selects executable browser actions while a separate text model handles planning.
- CDP Browser Automation with Jev
jev-browser-bridge is a CDP browser library that uses batched Jev Choice questions to select operations and DOM-derived targets.