Anthropic Adjusts Claude Max Subscription: Removes 5-Hour Limit and Renames Plans

Anthropic renames Claude Max subscription plans and removes the 5-hour usage limit for Max20, sparking positive community discussion.
Anthropic recently adjusted its Claude Max subscription plans, renaming Max5 and Max20 and removing the 5-hour usage limit under Max20. Previously, Max20 users triggered a quota reset every 5 hours; after removal, heavy users can use the service continuously without worrying about interruptions, significantly improving the experience for long coding, writing, or research sessions. The Reddit community expressed welcome with "yay," but the original information did not specify whether Max5 was adjusted simultaneously or mention price or other condition changes. Since the source is only a community post, official announcements have not been released; the specific new names, prices, and whether fair use policies apply still need to be based on official information. Overall, this adjustment reflects Anthropic's response to user feedback and may improve paid user satisfaction and retention.
Anthropic recently adjusted its Claude Max subscription plans, renaming the existing Max5 and Max20 plans and removing the 5-hour usage limit under the Max20 plan. This change sparked discussion in the Reddit community (see image below).

Subscription Plan Renaming and Limit Adjustment
According to information from the Reddit community, Anthropic has renamed the Max5 and Max20 plans. Although the specific new names were not clarified in the original information, this adjustment means users may see different labels at the subscription tier level. The more critical change is that the Max20 plan no longer has a 5-hour usage limit. Previously, Max20 users may have faced usage quota constraints that reset every 5 hours; with this limit removed, heavy users can use the service continuously over longer periods without worrying about hitting the limit midway.
Claude Max Subscription Plan Background
Claude is a large language model assistant developed by Anthropic, and its subscription system mainly includes the free version, Claude Pro (for general users), and Claude Max (for high-intensity usage scenarios). Claude Max was launched in 2024, positioned to meet the needs of heavy users such as developers and researchers, offering higher usage quotas and priority access. Initially, Max was divided into two tiers: Max5 (about $100 per month) and Max20 (about $200 per month), with the numbers roughly corresponding to usage multiples relative to the Pro plan. The so-called "5-hour limit" refers to the cap on the number of messages that can be sent within a rolling 5-hour window; once the cap is reached, users must wait for the window to reset. This type of rate-limiting mechanism is common in AI subscription services, used to balance server load and user experience.
Practical Impact on Users
For users who rely on Claude for long coding, writing, or research sessions, removing the 5-hour limit is a significant experience improvement. As a higher-tier subscription, Max20 was originally intended to meet high-intensity usage needs, but the 5-hour limit could interrupt continuous workflows. After the adjustment, users can schedule their usage more flexibly, especially suited for scenarios requiring work on large projects or deep conversations. However, the original information did not mention whether the Max5 plan also has similar limit changes, so Max5 users may still need to pay attention to its usage rules.
Usage Limits and Fair Use Policy
"Usage limits" in AI subscription services usually exist in two forms: first, hard quotas (such as N messages per 5 hours); second, fair use policies, which do not disclose specific values but reserve the right to rate-limit under abnormally high-frequency usage. Removing the 5-hour hard limit does not mean unlimited usage; service providers often still use fair use policies or dynamic rate limiting to prevent abuse. For Max20 users, the actual experience after the limit relaxation depends on backend capacity and scheduling strategies; during peak hours, even without explicit limits, response speed or availability may still be affected. Therefore, when understanding "removing limits," one should distinguish between "removing explicit quotas" and "completely unrestricted."
Community Reaction and Follow-up Observations
The Reddit post expressed welcome for this change with "yay," reflecting some users' positive attitude toward the limit relaxation. However, since the information source is only a community post, Anthropic has not yet released a detailed announcement. The specific plan names after renaming, whether prices will be adjusted, and whether the limit removal comes with other conditions (such as fair use policies) still need to be based on official information. Users are advised to follow Anthropic's official channels for accurate details.
Overall, this adjustment reflects Anthropic's response to user feedback, especially making concessions on usage limits, which may help improve paid users' satisfaction and retention. For users considering subscribing to Max20, removing the 5-hour limit increases the plan's appeal, but it is recommended to evaluate comprehensively based on one's own usage frequency and official terms.
Related articles

Prototype First, Then Let the Agent Build: A New Approach to AI Development
This article explores the practice of prototype verification (spike) before letting an Agent build in AI development, emphasizing experimental comparison and visualization for efficient human-machine collaboration.

Distill First, Then Fine-Tune: Get-Up Controller Success Rate Jumps from 71% to 94%
A developer trained a 2D ragdoll get-up controller in MuJoCo: end-to-end training from scratch failed completely; splitting into four policies plus a rule-based controller reached 71% then stalled; DAgger distillation plus PPO fine-tuning jumped to 94.2%. This article analyzes the break-down, distill, then end-to-end fine-tune approach, its anti-cheating validation, and limitations.

A Must-Read Safety Guide Before Using Claude Code to Clean Up Bloated TV Apps
A safety guide for using Claude Code to clean up preinstalled smart TV apps: manual permission mode, backups, disabling instead of uninstalling, a never-touch list, and closing debug ports to avoid bricking your TV.