I'm currently on the Max x5 plan and absolutely love using Claude Code, but it makes me really anxious when I'm not sure how much I have left before I hit my limit. I usually start getting nervous after around 10 prompts because I have no idea how many tokens I've used or how much time is left. I noticed that the token count can range from 2k to 10k, which adds to my anxiety. Why isn't there a progress bar or some sort of tracker to help us know how much we have left before reaching our limit? I only get a warning when I'm almost out, which is quite stressful. I know I could upgrade to the Max x20 plan, but it's a bit pricey for me right now. Any thoughts?
6 Answers
I've been using the Max plan for three weeks and haven’t seen any limit notifications! But if you’re doing steady coding, like refactoring, then you might run into that limit. If you pause often to think or test, you probably won’t hit it.
I've used the Max x5 for long coding sessions and haven't run into limits, even after coding nonstop for five hours. If you want to get more out of your usage, switching to Sonnet with the /model command is a solid option. They haven’t added a progress bar since usage can fluctuate based on Claude's capacity, which can be a bit unpredictable.
Honestly, it's tough to hit your limit if you're working with one agent. You’d have to keep executing tasks for a long time, like refactoring, to actually reach it. If you program a lot, it would be cool if they sold additional sessions once you hit your 50 limit because I probably hit that within 20 days each month!
I totally get where you're coming from! It would be super helpful if Anthropic added a progress tracking feature soon. But just so you know, if you do reach your limit, it’s not the end of the world. The limit is based on per session, which lasts for 5 hours. You do only get 50 sessions per month, so that can feel restrictive. If you're on the x5 plan, consider switching to the Sonnet model to increase your usage. Type /model to change it up!
Have you tried using /status? I thought it might help track things, but it only shows the model type you’re using, like Opus or Sonnet for your current session limits, which isn’t super useful if you ask me.
Yeah, it just tells you whether you're using Opus or Sonnet, not much else unfortunately.
I’m on the Max plan too, and I’ve noticed the output quality dips after heavy use. It makes fine-tuning tricky, and I can’t stand that! Do you face similar issues?
That’s actually because when you hit your quota for Opus, it switches you to Sonnet automatically, which has separate limits. So, longer chats might result in quicker downgrades to Sonnet, that’s likely what you’re seeing.
Yeah, I’ve only hit the limit when working non-stop on complex tasks, but if you take breaks, it’s easier to manage your usage.