gpt-6VIDEO
Sol vs Astra on a robot-drawing benchmark, tested
The author compared Sol and Astra on a "robot drawing on the board" benchmark, finding Sol's controller drew twice as fast and cost nearly 5x less, though spatial reasoning fell short of expectations.
Dmytro Hrybov
@dimentaryArchived 2026.09.24Originally posted 2026.09.23
Dmytro Hrybov @dimentaryOriginally posted 2026.09.23
compared GPT-6 Sol and Astra on my "robot drawing on the board" bench Sol used ~25k output tokens vs ~22k for Astra. interestingly, Sol's controller drew twice as fast to be honest, i expected stronger spatial reasoning from new Sol here, but it was way cheaper to run (almost 5x)
x.com/dimentary/status/2102539444834402314Archived 2026.09.24 10:10