Wanted to figure out which AI was best at hacking websites, so I put 11 LLMs against 32 challenges. GPT-5 solved 29/32 at 63% cheaper than Sonnet 4.5.
Big thanks to @SorceryIE for donating OpenRouter credits.
Blog: https://t.co/L3AoQHMRfn
Full Results: https://t.co/WkcyNCfF5l
We had an absolutely EPIC product launch, and to say THANK YOU we are offering an equally epic Cyber Week offer! Domain Data API for $1 a month! Want some? Just shoot us a message at [email protected] this week!