A developer has built a picture quiz game around TypeSafe's new Jev model, which has vision capabilities. Jev views a picture and a question, then picks an answer or taps a location in about a quarter of a second. The creator says the game was built to explore both the model's abilities and its speed, and the cards were made from scratch. Jev sees exactly the same picture the player does, and more cards will be added over time.
Players can compete against Jev alone, with each mode featuring its own leaderboard, or start a live match and challenge friends with Jev as one more participant. The project is open source, so others are invited to build their own versions or submit new cards for publication. The newest cards remain private for a period before being released.
The creator disclosed that Jev was run through both TypeSafe and Opper, and that the cards were produced using Opper's vision models and Claude Code. The author works at Opper, though the game can be played and run without using any Opper products.
The post appeared on Hacker News, where it received six points and one comment at the time of capture.
Source: Hacker News · Summarized by HeadlinesBriefing