... more token-heavy than 3.5. I d love to hear if anyone else has tried Claude 3.7 yet or has insights into these trade-offs. Let s discuss! Comment from Christian Safka(@christian_safka): From testing 3.7 vs 3.5 in Cursor Agent, I've found it's a lot better at estimating when it has enough context to implement or it needs to go check more files first. Looking forward to trying it some more Comment from Luke (@luke_arwayda ... ... Comment from Victoria Wu(@victoria_wu): @timoteolapa  Yeah. It's super cool! Comment from Victoria Wu(@victoria_wu): @andres_vlaeminck  It s interesting to hear that you found the 3.5 version to be more proactive in offering alternative