Real data: Claude Max weekly limit down 19%, and "Max 20x" is really Max 10x
Zapador · reddit · 2026-09-16
A Reddit user answered two hot questions with actual numbers from their own account: how much the Claude Max weekly limit really dropped, and what Max 5x vs 20x actually buys.
Key findings:
- The weekly cut is real and matches Anthropic's announcement (17%): measured drop is 19% (range 10–25%), from +50% promo to +25% of normal — and it took effect at the weekly reset, not on Sep 13 as announced.
- Max 20x is not 4× of 5x. Same account: the 5-hour session limit rose 4.4×, but the weekly limit only 2.2×. On 5x the weekly bucket holds 11 full sessions; on 20x only 5 — a more honest name would be "Max 10x".
- Dollar equivalents at API list prices: Max 5x ≈ $790/week, $70/session; Max 20x ≈ $1,710/week, $310/session (promo era).
- Moving to Fable 5.1 makes API-equivalent quota look halved for a separate reason.
Method: every Claude Code transcript parsed into a SQLite ledger, /usage endpoint polled every 2 minutes, with the never-promo'd 5-hour session limit used as reference to cancel distortions (cross-check: 11.2 vs 11.1 weekly/session ratio two months apart). Tooling open-sourced as ccstats.
Related event: Claude Max Weekly Limits Cut 19% in Real-World Tests(2 posts)→
More from Models
- Open-source VoxLock tops Krisp voice isolation benchmark with 4.95% WER — Arbaz Siddiqui · 2026-11-02
- Anthropic's 85% inference margin sparks debate on open source's price-pressuring role — code_star · 2026-11-02
- User reports xAI appears to have quietly raised weekly usage limits, dropping his meter from 93% to 47% — XFreeze · 2026-11-02
- Meituan's LongCat-2.5 lands as 1.6T MoE agent model, with DeepSeek copycat speculation — teortaxesTex · 2026-11-02
- Asked what AI firms want you to stop owning, Claude said: your own compute — alex_verem · 2026-11-02
- Is Hermes Still Worth It? Codex's $200 Tier Now Overlaps Most of Its Value — Foreign_Lead_3582 · 2026-11-02