I asked Jev to play N Wordle boards at once!

I asked Jev to play N Wordle boards at once! Each board hides a different word, but every guess applies to all unsolved boards simultaneously. Every move has to balance solving one board with revealing useful information across the others. Jev only needed 7 model calls across the entire game. Its shared guesses progressively narrowed the candidate set for every board until each remaining board had only one possible answer. From there, no more model calls were needed and Jev simply submitted the known answers. It cleared all 32 boards in 35 guesses and 5.4 seconds. The estimated API cost for the whole run was just $0.00068! I think Jev is really powerful for this kind of real-time decision loop. It is fast, extremely cheap, and can make one structured decision that advances many changing states at once. Excited to see what else can be built with these new models! https://x.com/richardcsuwandi/status/2100603807894053252?s=20