Anthropic Removes Claude Code Feature from New Pro Subscription, Shifts Claude Code to Higher-Priced Subscription
Multiple English-speaking users have reported and product page updates indicate that Anthropic has removed the Claude Code feature from the new Pro subscription. This means that new users who wish to use AI programming capabilities may need to switch to a higher-priced subscription plan.
Industry observations and developer discussions suggest that this move is related to the rapid growth in demand for AI programming. Claude Code is seen as a high-cost, high-computational feature, and similar strategies have previously appeared on other AI platforms, where high-value productivity tools are separated from general subscriptions and priced individually or with higher thresholds.
Source: Public Information
ABAB AI Insight
This adjustment essentially marks a shift in AI business models from "universal subscriptions" to "tiered pricing." Early models acquired user scale through low prices or even subsidies, but as reasoning costs and specialized scenario demands rise, platforms are beginning to price high-consumption, high-value features separately to restore unit economic models.
AI programming is one of the most typical "high value + high cost" scenarios. Compared to ordinary conversations, code generation involves longer contexts, higher call frequencies, and more complex reasoning paths, resulting in significantly higher marginal costs. At the same time, these users often directly convert AI into productivity tools and have a stronger willingness to pay, making price tiering an inevitable choice.
From a competitive structure perspective, this indicates that AI companies are starting to distinguish between "traffic users" and "professional users." The former is used to expand the ecosystem and data feedback, while the latter is the main source of revenue. The separation of Claude Code signifies that Anthropic is clearly positioning the developer community as core customers with high ARPU (average revenue per user), rather than as ancillary feature users.
In the long term, this pricing differentiation will drive AI capabilities towards an evolution of "free basic layer + paid production layer," similar to the early path of cloud computing. What truly determines the competitive landscape is no longer just the model's capabilities, but who can control costs while locking in those high-end usage scenarios that directly create economic value.