Abnormal token consumption rate on Opus 4.5 — 100% usage in ~1 hour
Status Closed — not planned
Maintainer reply None cached
Workaround ✓ Mentioned in thread ↓
Activity 13 comments · opened Mar 27, 2026 · closed May 12, 2026
Bug Report
Description
Token consumption has become drastically higher than normal, inconsistent with months of prior Claude Code usage on the same plan.
Observed Behavior
- This morning: Burned through 100% of usage allocation in approximately 1 hour
- Current session: Two light/low-activity sessions running simultaneously on Opus 4.5, already at 20% usage after only ~30 minutes
- Each session individually shows 10-15% usage despite minimal activity
Expected Behavior
Based on months of consistent Claude Code usage, this level of consumption is completely abnormal. Light sessions like these would historically use a small fraction of the daily allocation over the same time period.
Environment
- Model: Claude Opus 4.5 (claude-opus-4-5-20250514)
- Platform: macOS (Darwin 25.3.0)
- Claude Code: CLI
Evidence
I have timestamped screenshots confirming the usage spike and can provide them if needed.
Additional Context
- This is not a change in my usage patterns — same types of sessions, same workflows
- The spike appears to have started recently (noticed today, 2026-03-27)
- No unusually large file reads, no runaway agent loops — just normal interactive usage
<img width="1388" height="774" alt="Image" src="https://github.com/user-attachments/assets/296f862b-bc97-4655-a861-199afbf99571" />
<img width="2410" height="706" alt="Image" src="https://github.com/user-attachments/assets/c86864ec-8c19-488c-9785-e32b32e8e2ef" />
13 Comments
Related: #38335 — same symptoms (abnormal consumption starting around March 23). Also appears related to #38239, #39938, #39940, #37436, and #39492.
This seems to be a widespread issue affecting multiple Max plan subscribers simultaneously.
Found 3 possible duplicate issues:
This issue will be automatically closed as a duplicate in 3 days.
🤖 Generated with Claude Code
<img width="1279" height="288" alt="Image" src="https://github.com/user-attachments/assets/03616e2d-bc88-4df0-81de-610f5d2d4103" />
self-sabotage.. I'm out..
Yeh, im hitting abnormal usage too...
Normally 1x 5 hour limit is 8% for my total week.. During promo it was 5-6%...
Today? Its 14 fking percent... Thats HALF my usage per week...
What the actual fk is going on here???????
Max20 here...
Using Max5 for a few tasks (creating pages, adding buttons, fixing flows) - normal usage ~4% of 5-hour, asked claude to move the button, boom 20%
<img width="1518" height="523" alt="Image" src="https://github.com/user-attachments/assets/0096f85e-e7f4-4b96-912c-769a0754dff9" />
Same pattern on Opus 4.6 now. Max 20, v2.1.89, April 1: 100% in ~70 min after reset.
Full report: #41788
Related: #38335, #38239, #40895, #23706, #41663
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_URLenv 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
<img width="583" height="360" alt="Image" src="https://github.com/user-attachments/assets/2cea7f11-1d83-4a8b-8301-7f9456da7905" />
Done in ~3 days and i slowed down the last day and i was on .88. and .89...
Zero difference for me... Thats about HALF my usage this week...
This is max20... 200 usd per month... Unacceptable....
Everyone should cancel their sub, even if you intend to resume it in x weeks? Cancel it... If 10-100k of us do it? It will be statistically big enough to force their hand... Its the only way at this point...
Follow-up — precautions (April 2, 2026)
100% in ~1 hour on Opus = consistent with 0% cache read (each turn rebills full context).
Additional drain accelerators:
--resume(token bomb),/dream&/insights(background drain), v2.1.89Ref: 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
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
Closing 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.