Salvatore Sanfilippo finds GPT-5.5 effective for DS4 coding

Redis creator Salvatore Sanfilippo reported that every participant in the DS4 project received substantial coding assistance from GPT-5.5. Multiple contributors who compared the models directly found that Anthropic's Claude 3 Opus produced no useful output on the same tasks. Engineer Adam.GPT reposted the account of the results.

Original post

T(#400@TEORTAXESTEX@ANTIREZ

AN

antirez@ANTIREZ

Gentle reminder on how, in the recent DS4 fiesta, not just me but every other contributor found GPT 5.5 able to help immensely and Opus completely useless.

2:19 AM · May 14, 2026 View on X