gpt-6POST
GPT-6 Astra feels dumber and burns quota faster, one user says
The poster says Astra has gotten noticeably less capable and uses up quota faster since launch, and now prefers Gemini 3.8 Flash for its generous limits and lack of token anxiety.
John
@johnAGI168Archived 2026.09.14Originally posted 2026.09.13
John @johnAGI168Originally posted 2026.09.13
GPT 6 Astra 神一样的登场，现在是越用越笨，额度消耗越用越快。降智来得有点快 反观 Gemini 3.8 flash，大家都没什么关注。现在我反而感觉 Gemini 3.8 flash 越用越好用，量大管饱随便用，不行改到行为止，又快又好用。没有 token 焦虑，就和手机没有续航焦虑是一个道理！
x.com/johnagi168/status/2099142190555369548Archived 2026.09.14 16:05