I've been using the new $100 Claude Max plan and I'm curious if others have experienced any changes in Claude Code's context handling. I was a fan of Claude Code, paying $3–$7 per task, but after switching to Max, I've noticed I'm getting more errors and the results seem less reliable. I tested a new feature for my React Native app that requires good context, and the outcomes weren't as impressive as they were with the API. I'm starting to think I might want to revert back to the yearly Pro plan. Has anyone had a similar experience or noticed any differences?
6 Answers
Without clear evidence, it’s hard for anyone to say for sure. You might need to conduct a few more tests to get solid results!
I haven't noticed any differences at all. To me, it feels the same using both the API and the subscription. Claude Code still impresses me!
I had the same concerns and even reached out to support. They mentioned there's a difference in throughput between API and Max, which could be affecting performance. I've written a couple of posts about my findings if you want to take a look.
I experienced something similar! I switched from the desktop version to the Max plan and the differences were striking. Everything felt off!
Interesting! I just upgraded yesterday and I’m kinda curious about what others have found too. Let me know if you find out anything more!
Honestly, I just tried the Max plan and find it way better! I love that I'm not stressing over API costs anymore, and Claude Code handles scripting and debugging really well now. It feels like a better workflow overall.
Yeah, but that doesn't really address the context issue the original poster mentioned.