Anthropic, the AI research and safety company, has officially restructured its premium subscription tiers, rebranding its high-end Claude offerings while removing restrictive usage windows that have long frustrated power users. In a Friday announcement posted via its official account on X, the company confirmed that the $100-per-month "Max 5x" plan is now simply titled "Max," while the $200-per-month "Max 20x" plan has been rebranded as "Ultra." This strategic realignment of nomenclature is accompanied by a significant functional update: verified Ultra subscribers are no longer subject to the rigid five-hour message capacity limits that previously throttled their workflow.
The Shift in Subscription Architecture
The modifications to the subscription model represent an effort by Anthropic to streamline its user experience and provide more predictability for professional developers and heavy AI users. Under the new framework, the Max plan remains priced at $100 monthly, offering five times the usage capacity of the standard Pro tier, along with $100 in monthly API credits. The Ultra tier, retailing at $200 per month, now provides ten times the usage capacity of the Pro tier and includes $200 in monthly API credits.
Perhaps the most impactful change for the user base is the removal of the five-hour rolling limit for verified Ultra accounts. Previously, even the most expensive tiers were subject to hard caps that reset every 300 minutes, a design intended to manage server load but often criticized by power users for interrupting complex, multi-stage tasks.
Anthropic clarified that these changes apply automatically to existing subscribers. However, the company has implemented a transitional verification period for newer accounts. To mitigate the risks of fraud and automated abuse—common issues for large language model (LLM) providers—new accounts will remain subject to temporary limits until they undergo verification. Once cleared, these users will be notified via email that their restrictions have been lifted. Weekly usage caps, however, remain in effect across all tiers to ensure continued stability of the underlying infrastructure.
A Chronology of Claude’s Evolving Service Tiers
The path to this current subscription model has been marked by a series of rapid adjustments as Anthropic seeks to balance infrastructure costs with the high demand for its models, specifically the flagship Fable series.
- April 2026: In an effort to maintain platform integrity, Anthropic took a firm stance against third-party integrations, banning tools like OpenClaw that attempted to wrap or automate Claude subscriptions. This move signaled the company’s intent to maintain a "walled garden" approach to its premium services.
- May 2026: Bolstered by increased computing capacity—famously facilitated by the deployment of SpaceX’s Colossus supercomputing cluster—Anthropic doubled the five-hour message limits for Claude Code, marking the first major easing of restrictions following the initial launch of their developer-centric tools.
- June 2026: The company faced legal headwinds when a lawsuit was filed challenging the advertising of "Max" plans, with plaintiffs alleging that the company oversold usage capacities and failed to meet the promised throughput levels.
- November 2026: The current announcement marks the latest iteration of this service strategy, moving away from the "5x" and "20x" terminology, which had become a point of contention and confusion for users.
Analyzing the Capacity Adjustment
While the rebranding simplifies the user interface, the reduction in the stated multiplier for the Ultra plan—from the former "20x" to the current "10x"—has drawn attention from industry analysts. By anchoring the new Ultra plan at ten times the Pro tier, Anthropic appears to be aligning its marketing language with the reality of its actual server performance. The previous "20x" label had created an expectation of volume that proved difficult to maintain during peak global usage hours.
The integration of Fable, Anthropic’s most advanced model, serves as a cornerstone of these premium offerings. Users on both the Max and Ultra plans now receive access to Fable for up to 50% of their total weekly allowance. This allocation ensures that the most computationally expensive model is prioritized for those willing to pay the premium, while still allowing the system to scale effectively across the broader user base.
For international customers, the pricing remains consistent with local currency adjustments. In European markets, the tiers are priced at €90 and €180 per month, respectively, reflecting Anthropic’s commitment to maintaining parity across its major operational regions despite fluctuations in exchange rates and operational costs.
Official Stance and User Experience
In its communication to the public, Anthropic emphasized that the transition should be seamless. "If you’re a subscriber, there’s nothing you need to do," the official Claude account stated. This "zero-friction" approach is critical for the company, as it attempts to retain enterprise and power-user loyalty amidst fierce competition from OpenAI’s ChatGPT, Google’s Gemini, and various open-source alternatives.
The removal of the five-hour cap is a direct response to feedback from the developer community. For a developer or a data analyst, being cut off mid-session is more than a minor inconvenience; it disrupts the "flow state" required for high-level coding or technical writing. By shifting the limitation to a weekly volume, Anthropic provides users with the flexibility to "burst" their usage during high-intensity periods without the penalty of being locked out for several hours.
Broader Implications for the AI Economy
The evolution of Anthropic’s pricing structure highlights a broader trend within the artificial intelligence sector: the transition from "experimental" AI pricing to "enterprise-grade" utility. Early iterations of AI subscriptions were often characterized by vague limits and unpredictable performance. As the market matures, companies are moving toward more transparent, usage-based, or capacity-guaranteed models.
The legal and operational challenges Anthropic faced in 2026 illustrate the inherent tension between the massive compute requirements of frontier models and the need to offer a consistent, reliable service. The collaboration with entities like SpaceX to expand computing capacity demonstrates that the bottleneck for AI companies is no longer just software innovation, but the physical infrastructure required to run inference at scale.
Furthermore, the focus on "verified accounts" underscores a growing concern regarding the security of premium AI services. As these tools become more capable, the potential for them to be used for mass-scale scraping, spam generation, or other malicious activities increases. By gating the removal of usage limits behind a verification process, Anthropic is effectively creating a tiered trust system, ensuring that their most powerful, uncapped resources are utilized by legitimate human users and verified organizations.
Future Outlook
As the industry moves into 2027, the focus for providers like Anthropic will likely shift from basic capacity management to value-added features. With the current restructuring, Anthropic has effectively "cleaned up" its product offerings, removing the ambiguity of the "20x" branding and providing a more reliable, if slightly more constrained, service environment.
Whether these changes will satisfy the plaintiffs in the ongoing June lawsuit remains to be seen. However, from a product management perspective, the move is a pragmatic step. By aligning their marketing with sustainable usage patterns, Anthropic is positioning itself to better manage customer expectations, reduce the frequency of service outages, and provide a more stable platform for the high-end developers and enterprises that form the backbone of their business model.
For the average subscriber, the message is clear: while the terminology has changed, the underlying goal is a more predictable, reliable, and accessible experience. As the competition between AI models intensifies, the companies that succeed will likely be those that can best balance the immense cost of compute with the demand for seamless, uninterrupted access to their most powerful AI agents. Anthropic’s latest update is a significant, albeit reactive, move in that ongoing struggle for market dominance and user trust.
