Claude Max 20x Usage Cap Applies Only to 5-Hour Window, Not Weekly Limit
Anthropic's Claude Max subscription tiers advertise "5x or 20x more usage than Pro," but the multiplier applies only to a 5-hour session window, not the weekly limit. A Hacker News post by user vmg12 highlights the confusion, pointing to a lawsuit alleging false advertising and a tweet that clarifies the actual terms.
Key Details from the Discussion
- The marketing copy on Anthropic's site shows "Choose 5x or 20x more usage than Pro*" without explicitly stating the 5-hour window in the headline.
- The FAQ buries the detail: "You get everything in Pro, plus more usage: choose from 5x or 20x the usage of Pro per 5-hour session."
- No comparison table on the site shows weekly limits for Pro, Max 5x, and Max 20x — making it hard to judge the real value.
- User
rybosworldnotes that "2 $100 plans provides more usage than 1 $200 plan" — meaning two 5x subscriptions may give more total weekly usage than one 20x subscription. - User
AregNoyaadds historical context: the 20x tier was introduced before weekly limits existed, and the 20x tier only offers roughly 1.7x the weekly usage of the 5x tier.
Practical Takeaway
If you're choosing between Claude Max plans, don't rely on the headline multiplier. Check the FAQ for the 5-hour session limits and estimate your weekly usage. For heavy daily use, two 5x accounts might give you more total capacity than a single 20x account — and at roughly the same price ($200 vs $200).
The lawsuit against Anthropic is ongoing, and the company has not yet updated its marketing copy. In the meantime, treat "20x" as a session-based boost, not a weekly cap increase.
📖 Read the full source: HN AI Agents
👀 See Also

Linux kernel maintainer reports sudden shift in AI-generated bug report quality
Greg Kroah-Hartman says AI-generated bug reports for the Linux kernel went from 'AI slop' to legitimate reports about a month ago, with open source security teams across projects seeing the same shift. The kernel team is handling the increase with tools like Sashiko for review automation.
Uber fined €825M by Dutch regulator over algorithm-driven driver deactivations
The Dutch data protection authority fined Uber €825 million for using an algorithm that automatically deactivated driver accounts without proper human oversight, violating GDPR.

Claude Code 2.1.132: Multiagent docs, schedule gate, skill limit changes
Release v2.1.132 adds Managed Agents documentation for multiagent sessions, outcomes, and webhooks; introduces a default-deny gate for proactive /schedule offers; drops documented skill limit from 64 to 20 per agent.

xAI loses legal challenge to California AI data disclosure law
xAI has lost its attempt to block California's AI data disclosure law, which requires companies to disclose training data sources and other details about their AI systems. The court ruling means the law will proceed as scheduled.