[BUG] 20x Max Plan User Hitting "Extra Usage" Limits After Single Session - Severe Regression
Status Closed — not planned
Maintainer reply None cached
Workaround ✓ Mentioned in thread ↓
Activity 11 comments · opened Mar 31, 2026 · closed May 12, 2026
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
As a 20x Max Plan subscriber, I'm hitting "extra usage" limits after only ONE chat session, with the limit resetting at a specific time (6pm Europe/Istanbul). Previously, I could run 10+ parallel Claude Code sessions simultaneously, complete multiple tasks, and never encounter any usage limits throughout the day.
This represents a severe regression in service quality for Max Plan subscribers who are paying premium prices for higher usage limits.
Key Observations:
- Only 155 tool uses across 9.5 minutes of work
- Session completed only ONE task before hitting limit
- Error mentions "extra usage" - implying there's a base usage + extra usage separation
- Limit resets at a specific time (6pm local time) - suggests daily quota system
- Previously could complete 10+ sessions with hundreds of tool uses each
What Should Happen?
Historical Performance (Same Account, Same Plan):
- Could run 10+ parallel Claude Code sessions simultaneously
- Each session could use **hundreds to thousands of tool uses
- Could work for hours without hitting any limits
- No "extra usage" concept existed or was never triggered
- No time-based reset mechanisms were visible
What Max Plan Should Provide:
- Unlimited or very high usage limits consistent with the premium tier pricing
- Ability to run 10+ parallel sessions as was previously possible and documented
- No artificial throttling that prevents normal development workflow
- Clear communication if limits have changed from previous behavior
- Accurate token tracking (current session shows "0 tokens" which seems incorrect)
Error Messages/Logs
Steps to Reproduce
- Prerequisites: Be a 20x Max Plan subscriber
- Start a Claude Code session with model
opus 4.6 default 1m context - Use the session normally:
- Make tool calls (file operations, code generation, etc.)
- Complete a moderate-sized task (~150 tool uses)
- Work for ~9-10 minutes
- Observe: After approximately 155 tool uses, receive error:
``
You're out of extra usage · resets 6pm (Europe/Istanbul)
- Attempt: Try to continue the session or start a new one
- Result: Blocked from continuing until the reset time (6pm local)
- Compare with previous behavior: Same account previously allowed 10+ parallel sessions with no such limits
Claude Model
None
Is this a regression?
Yes, this worked in a previous version
Last Working Version
_No response_
Claude Code Version
2.1.88
Platform
Anthropic API
Operating System
macOS
Terminal/Shell
Terminal.app (macOS)
Additional Information
_No response_
11 Comments
Found 3 possible duplicate issues:
This issue will be automatically closed as a duplicate in 3 days.
🤖 Generated with Claude Code
They reduced the quota. Literally. It's not a bug, technically. Now they mark all reports as "invalid".
Well there goes my subscription then , I am cancelling it for OpenAI I guess.
Experiencing the same on Max 20 ($200/mo) — rate limit 100% exhausted in ~70 minutes with Opus.
Root cause I identified via a monitoring proxy (
ANTHROPIC_BASE_URL): two client-side cache bugs (#40524, #34629) cause the cache read ratio to drop to ~4%, inflating token consumption by ~20x.Downgrading to v2.1.68 restored cache read ratio to 89–99%. @Commandershadow9 independently verified this with A/B testing across 21 sessions (#41506).
Full analysis + safe workarounds: https://github.com/ArkNill/claude-code-cache-analysis
Note: Tested in CLI (
claudein terminal, Linux). If you're using the VSCode extension, downgrading the extension separately to v2.1.68 has also been reported to help (#38335).Follow-up — precautions (April 2, 2026)
"Extra Usage" limits after a single session on 20x = severe cache miss inflating every turn.
Additional behaviors to avoid:
--resume— replays full history as billable input (token bomb)/dream,/insights— hidden background token drainRecommended: v2.1.81 (fixed) + fresh sessions + proxy monitoring.
Ref: https://github.com/ArkNill/claude-code-cache-analysis
Update (April 2): v2.1.90 has significantly improved cache efficiency — benchmark shows 95-99% cache read in stable sessions (both npm and standalone installations).
If you're still affected:
claude update(ornpm install -g @anthropic-ai/claude-code)"DISABLE_AUTOUPDATER": "1"to~/.claude/settings.jsonenv section--resume(still broken)Note: server-side quota issues (org-level pool sharing, accounting mismatches) remain unresolved — the above fixes the client-side cache drain only.
Benchmark data: https://github.com/ArkNill/claude-code-cache-analysis
Thats the biggest bullshit ever... and they are not able to fix it asap... Howmany "seniors" are developing for them? 1?
Thats a absolute desaster, my #42678 was working yesterday also my #42681 was working before but now all is going to hell.
I am also experiencing a similar issue. I only ran a single simple query, and after checking /context, it showed that 61k tokens were used, yet it already consumed 10% of the 5-hour session limit.
Unfortunately, I just renewed two accounts earlier this week, costing $440.
Brief update (April 3): v2.1.91 is out and cache behavior is further improved — npm and standalone now perform identically (Sentinel gap closed). However, testing revealed two additional unfixed bugs: a 200K tool result budget cap that silently truncates older results, and a client-side false rate limiter generating
<synthetic>errors without API calls (151 entries confirmed). If you are still seeing drain after updating, these are the likely causes. Details: claude-code-cache-analysisClosing for now — inactive for too long. Please open a new issue if this is still relevant.
This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.