Pro Max 5x quota exhausted in 1.5 hours despite moderate usage
![[BUG] Pro Max 5x Quota Exhausted in 1.5 Hours Despite Moderate Usage · Issue #45756 · anthropics/claude-code](https://pub-90f0ac00d93c47daac3e1d2cdd28d496.r2.dev/articles/60c184b89f022da6788b05f0fba00ee6.webp)
github.com
April 12, 2026
6 min read
72/100
Summary
The Pro Max 5x plan for Claude Code has reportedly exhausted its quota in just 1.5 hours despite moderate usage. Users are experiencing unexpected quota depletion issues.
Key Takeaways
- The Pro Max 5x plan's quota was exhausted in 1.5 hours despite moderate usage, raising concerns about token accounting.
- Cache read tokens appear to count at full rate against the rate limit, negating the benefits of prompt caching for quota purposes.
- During heavy development, the system consumed 24.4M effective tokens per hour, while moderate usage led to an unexpected consumption of 70.5M tokens per hour.
- Each API call sends the full context as input, resulting in significant quota consumption, especially when cache read tokens are not discounted as expected.
Coverage & discussions
Hacker News · github.com
Community discussions
Community Sentiment
Positives
- The recent UX improvements to prompt caching are a step in the right direction, potentially reducing the frequency of costly cache misses during long sessions.
- Users report that switching to the Codex plan has resulted in a more generous quota and improved accuracy for their specific tasks, indicating better performance in certain use cases.
- Despite some issues, the Codex model is perceived as a better deal compared to Claude Code, suggesting a competitive edge in terms of quota management.
Concerns
- Many users are experiencing significant performance regressions with Claude Code, including prolonged exploration loops that lead to rapid quota exhaustion.
- The recent change in prompt caching behavior has left users frustrated, as it appears to negatively impact token usage and overall efficiency.
- There are concerns that the quota limits are becoming increasingly restrictive, making it difficult for users to utilize the service effectively for their work.
Related Articles

Anthropic silently downgraded cache TTL from 1h → 5M on March 6th
Apr 12, 2026
Excessive token usage in Claude Code
Feb 21, 2026

Claude Code Just Got 5x More Expensive
Jun 30, 2026

Claudetop – htop for Claude Code sessions (see your AI spend in real-time)
Mar 14, 2026

Claude Code users hitting usage limits 'way faster than expected'
Mar 31, 2026