lot's of availability issues with @AnthropicAI today.
it sucks because you agent stop working and you only notice when you get back.
claude code could proactively recover from this kind of error and continue the session @trq212
Token smartness is the biggest chalenge we still have.
Look at the numbers of the API cost for my usage in codex and claude code this week. It's doing a great job, but the aggregate value it generated is far from cost.
PS: Had two codex reset this week.
@OpenAI@AnthropicAI
@icanvardar But, according to benchmarks, opus 5 high matches inteligente with codex and costs less. Check it out. (inteligence 59 vs 59)
I love codex, it was my first AI coding tool - so I have feeling for codex. But my current experience with limits using PRO plan hasn't been great.
@ArtificialAnlys@AnthropicAI There is something weird happening here.
In artificial analisys opus 5 reaches lower deepswe (63?) than deepswe site tells (74). What is going on???
You know what?
Right now the subscription that is the most worthful in the entire world is the Claude one. With Opus 5 the $200 plan will look like infinte - and now they have the best model.
This right now. Next week? Who knows?