Arturo

🔥 Building "Road2Crypto" web/iPhone app.

[Arturo @MrCrypto305]

@goodworse All I’m seeing right now is degraded reasoning from 5.5, it feels like 5.3. I switched back to opus 4.8, wake me up when it’s fixed

[Arturo @MrCrypto305]

@sethrosen 😂

[Arturo @MrCrypto305]

Claude Opus 4.8 max is way ahead of Codex, I love codex but until they fix it I’m switching to claude, gpt 5.5 right now feels like 5.3

[Arturo @MrCrypto305]

@nejatian Very good!

[Arturo @MrCrypto305]

@messedupfoods I see that and I turn around.

[Arturo @MrCrypto305]

@Yoda4ever I just want to know how it ends 😂 how will it stop it now hahah

[Arturo @MrCrypto305]

@ValsAI It should be 0 then not a fallback

[Arturo @MrCrypto305]

@MuzanKizuki @AnimexTwts Who’s him?

[Arturo @MrCrypto305]

@Ananth7e You are kidding right? 5.5 makes such a slop code that looks like gpt3

[Arturo @MrCrypto305]

@imjustnewatai Where dod you get that graph from? Trust me bro?

[Arturo @MrCrypto305]

@ClaudeDevs Nice!

[Arturo @MrCrypto305]

@iruletheworldmo That’s an insane jump

[Arturo @MrCrypto305]

@cognition Excellent

[Arturo @MrCrypto305]

@CFchangelog Nice!

[Arturo @MrCrypto305]

@auroter Completely agree, I paused my coding until the 5.6 release so it can become smarter again and when it does I’ll setup a local model in codex. This is simply unacceptable. You get from work to work on extra projects and endup “i don’t have energy to deal with codex stupidity”

[Arturo @MrCrypto305]

@TeksEdge Forgot to include Qwen in the benchmark…

[Arturo @MrCrypto305]

They have nerfed ChatGPT 5.5 so much… you can tell a new release is coming so people feel it’s “smarter” these companies are full of scams. Once we get a 5.5 level of a local llm they are cooked

[Arturo @MrCrypto305]

@hthieblot Jobs

[Arturo @MrCrypto305]

@jpschroeder I hope not

[Arturo @MrCrypto305]

@WinRAR_RARLAB 🤣