WHAT NEW IN X10.28 WED

AI collections from X · 22 collections · 58,924 posts

jevIMAGE

Browser and computer use tasks: Luna 96.5% vs Jev 37.3%

In a test of basic browser workflows, more real-world end-to-end tasks and some computer use tasks, Luna succeeded 96.5% of the time overall while the best Jev implementation found succeeded only 37.3%.

Steve (Builder.io)
@Steve8708Archived 2026.09.24Originally posted 2026.09.23
Browser and computer use tasks: Luna 96.5% vs Jev 37.3%
Steve (Builder.io) @Steve8708Originally posted 2026.09.23
Open original on X

I ran a set of basic browser workflows, more real-world end to end tasks, and some computer use tasks Luna was able to accomplish 96.5% overall, while the best Jev implementation we found succeeded only 37.3% of the time

x.com/steve8708/status/2102534746110906661Archived 2026.09.24 10:19