[BUG] Not even doing real work and we're at 40% Fix the rate limits (Max x5)

Found 3 possible duplicate issues:

1. https://github.com/anthropics/claude-code/issues/38350
2. https://github.com/anthropics/claude-code/issues/41504
3. https://github.com/anthropics/claude-code/issues/41035

This issue will be automatically closed as a duplicate in 3 days.

• If your issue is a duplicate, please close it and 👍 the existing issue instead
• To prevent auto-closure, add a comment or 👎 this comment

🤖 Generated with Claude Code

Claude is just as inefficient (I'd say even more so than a week ago) if it consumes the previous in-chat context on resume, it's worthless to us, take it off or make it an option.

The second prompt that got me 20% added on top of the other 20%, it just gave me a command and didn't even run it. So supposedly, it didn't even consume the previous context.
The only thing left for us to do is start collecting community funds to sue you.

Found the fix! Refund Claude and buy a ChatGPT subscription.

Same. Max 20 ($200/mo), v2.1.89. April 1: 100% in ~70 min after reset without doing real work either. Rate limits are fundamentally broken.

Full report: #41788
Related: #38335, #38239, #41663, #41506, #41084

I've been experiencing the same issue on Max 20 ($200/mo) — rate limit 100% exhausted in ~70 minutes.

After setting up a monitoring proxy using the official ANTHROPIC_BASE_URL env var, I identified two cache bugs as the root cause (#40524, #34629) and measured the impact: cache read ratio dropped to 4.3%, meaning ~20x token inflation per turn. After applying workarounds it stabilized at 89-99%.

Full analysis with per-request measured data, safe workarounds, and community references (including cc-cache-fix): https://github.com/ArkNill/claude-code-cache-analysis

Update (April 2): Client-side cache bugs are largely fixed in v2.1.90. However, if you're hitting limits with zero/minimal actual usage, that points to a server-side accounting issue or org-level quota sharing (accounts on the same billing share a pool — #41881).

For the client-side fix: update to v2.1.90 + disable auto-update. Details: https://github.com/ArkNill/claude-code-cache-analysis

Update (April 2): Client-side cache bugs are largely fixed in v2.1.90. However, if you're hitting limits with zero/minimal actual usage, that points to a server-side accounting issue or org-level quota sharing (accounts on the same billing share a pool — #41881). For the client-side fix: update to v2.1.90 + disable auto-update. Details: https://github.com/ArkNill/claude-code-cache-analysis

Thanks bro, I already decided to move forward with ChatGPT. Thanks for the attention again.

April 3 update: v2.1.91 fixes the cache regression that caused the worst drain. If you are still hitting limits after updating, there are additional unfixed mechanisms: a 200K tool result budget cap, a client-side false rate limiter, and silent context stripping — all confirmed via proxy testing. Anthropic acknowledged peak-hour tightening on X (Lydia Hallie) but stated "none were over-charging you." Measured data and analysis: claude-code-cache-analysis

April 3 update: v2.1.91 fixes the cache regression that caused the worst drain. If you are still hitting limits after updating, there are additional unfixed mechanisms: a 200K tool result budget cap, a client-side false rate limiter, and silent context stripping — all confirmed via proxy testing. Anthropic acknowledged peak-hour tightening on X (Lydia Hallie) but stated "none were over-charging you." Measured data and analysis: claude-code-cache-analysis

Thanks for the update, I'll be rocking GPT-5.4 for the time being, only 258k context on Codex but as much usage as I can virtually pump out.

My Claude says if you're on v2.1.81, a known cache bug in v2.1.87-89 and a client-side rate limiter issue (#40584) may be contributing to your usage depletion. Issue #41506 has the most comprehensive analysis from two users who measured this from both client-side and server-side — including a timeline of which versions are affected and which fixes help. Updating to v2.1.91+ resolved the cache drain portion for multiple users, though some report capacity still feels tighter than before March 28.

My Claude says if you're on v2.1.81, a known cache bug in v2.1.87-89 and a client-side rate limiter issue (#40584) may be contributing to your usage depletion. Issue #41506 has the most comprehensive analysis from two users who measured this from both client-side and server-side — including a timeline of which versions are affected and which fixes help. Updating to v2.1.91+ resolved the cache drain portion for multiple users, though some report capacity still feels tighter than before March 28.

https://aistupidlevel.info/

I'm using GPT for now, and I'm very much satisfied.

40% usage without doing real work means something in the session is consuming tokens silently — usually old tool outputs and metadata being re-sent every turn. Cozempic's guard daemon prunes this bloat continuously with 18 strategies so your token budget goes to actual work. pip install cozempic — hooks wire on first run. https://github.com/Ruya-AI/cozempic — happy to hear how it goes.

Closing for now — inactive for too long. Please open a new issue if this is still relevant.