@cremieuxrecueil I'm fully using this mode. It's one-shotting all the tasks, and it's really impressive. I even tried a vague task and did really well! I have the $200 sub, and I'm not seeing any problem for now.
@Mohsine_Mahzi@mark_k@AnthropicAI The Claude app has a lot of bugs. The ChatGPT web app is way better than the app as well. I'd say that doesn't affect the final user this way, specially because most of the companies need IT approval to download an app
@shiri_shh@AnthropicAI Mythos is the only way Anthropic can keep on track against OpenAI's publicly available models, especially if OpenAI drops 5.6 today
@Mohsine_Mahzi@mark_k@AnthropicAI I don't think that's over just because Windows user don't have some features... CLI is available anyway. And we can't even compare the quality of code between Opus 4.7 and GPT 5.5.... Rumor has it that GPT 5.6 will beat any version of Opus 4.8; let's see
@Angaisb_ Opus 4.8 vs. GPT 5.6. Which should we try first? Users with pre-access shared that if Anthropic doesn't release Mythos, GPT 5.6 will kill Opus.
@leanderriefel Yep, that's true! I was running 3 goals at the same time in some big projects, hit my weekly limit 2x (one before the reset and about to hit another one now). That's not OpenAI's fault.