[BUG] Paid Max 5x, weekly limit still enforced at Pro size (2nd paid week, support escalation unanswered for 5 days)

Status Open
Reported on v2.1.283
Maintainer reply None cached
Activity 0 comments · opened Sep 26, 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?

Upgraded from Pro to Max 5x on 21 September 2026. Billing went through and claude.ai Settings > Usage shows "Plan usage limits: Max (5x)". The weekly limit is still enforced at Pro size. This is now the second paid weekly window on Pro limits.

Measured from local Claude Code transcripts (input + cache writes + output, cache reads excluded), compared with the weekly meter:

| Weekly window | Plan paid | Tokens per 1% of week |
|---|---|---|
| from 5 Sep | Pro | ~612,000 |
| from 19 Sep | Max 5x (from 21 Sep) | ~764,000 |
| from 26 Sep | Max 5x | ~840,000 |

With the Max 5x multiplier, 1% should be about 3,000,000 tokens.

Example, 26 Sep 2026, 10:00 to 10:55 UTC (right after the weekly reset): ~5.9M tokens moved the weekly meter to 7%. On Max 5x that should be ~2%.

The usage data inside the Claude Code desktop app still reports plan: "Pro".

What Should Happen?

The Max 5x weekly limit is applied to the account, matching what I pay for and what claude.ai Settings shows. The paid days that ran on Pro limits are compensated.

Error Messages/Logs

Steps to Reproduce

  1. Be on Pro, upgrade to Max 5x mid-cycle (21 Sep 2026).
  2. Wait for the next weekly reset (26 Sep 2026, 10:00 UTC).
  3. Use Claude Code normally and compare tokens used with the weekly meter.
  4. The meter still moves at Pro speed. The usage data still reports plan "Pro".

Claude Model

Opus

Is this a regression?

No, this never worked

Last Working Version

_No response_

Claude Code Version

2.1.283 (Claude Code)

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

Other

Additional Information

Support: conversation ID 215476025410347. On 21 Sep the support agent wrote that ourmethodology is sound, that the enforcement system is not applying the Max 5x multiplier, and that the window starting 26 Sep should run on Max 5x. It was handed to the human team on 21 Sep. No human reply after 5 days. A follow-up on 26 Sep got only an automated "you are already in our queue".

Same pattern as #58101, #76093 and #79773.

Using the Claude Code desktop app on Windows 11, model Claude Opus 5.5.

View original on GitHub ↗