[BUG] Claude 4 (or since its release) is BURNING USAGE LIMITS x10 TIMES FASTER

Environment

• Platform (select one):
• [ ] Anthropic API
• [ ] AWS Bedrock
• [ ] Google Vertex AI
• [X] Other: Dedicated server under LXC machine
• Claude CLI version: Claude Code v1.0.2
• Operating System: Ubuntu 25.04 plucky
• Terminal: Putty SSH

Bug Description

Before last Claude 4 release worning under claude 3.7, code usage was working perfectly, very difficult to reach limits even working 2 people under same machine using claude code for a full long day of work managing hundreds of files codding.

Since last update with Claude 4 I am reaching the limit only by myself working on little files managing code in less than 3 hours and having to wait almost 3 hours to continue working.

This is unacceptable viewing the cost of MAX plan. This might be 2 possible situations.

1) Claude's 4 AI update is Greedy as hell pumping tokens and you didn't adapted MAX limits to new Claude's behaviour and needs to be fixed URGENTLY.

2) You are 100% conscient of new Limits and Anthropic is abusing on its client's knowing that we are actually paying a Super Premium Max Plan. I want to beleave you are not so Greedy and that's not your strategy to face Google's new AIs.

Steps to Reproduce

1. Ask ANYONE how many hours and deep working needed to reach MAX limits.
2. Ask now to any programmer how are being feeling those "new" limits
3. make your own limits testing comparing both models/compilations

Expected Behavior

To work with the exact same limits than before Claude 4 release.

Actual Behavior

Screamming over my screen angry as hell prarying to be fixed ASAP