Back to List
Anthropic to Restrict Claude Code Usage with Third-Party Tools Due to Subscription Design Constraints
Industry NewsAnthropicClaude CodeAI Subscriptions

Anthropic to Restrict Claude Code Usage with Third-Party Tools Due to Subscription Design Constraints

Anthropic has announced plans to restrict the use of Claude Code when integrated with third-party tools and harnesses. The decision was communicated by Boris Cherny, the head of Claude Code, via a statement on X (formerly Twitter). According to Cherny, the current subscription models for Claude Code were not originally designed to accommodate the specific usage patterns generated by external third-party harnesses. This move highlights a strategic shift in how Anthropic manages its developer tools and subscription structures, ensuring that usage remains aligned with the intended design of their service tiers. The restriction aims to address discrepancies between user behavior on third-party platforms and the underlying subscription framework provided by Anthropic.

Tech in Asia
Share

Key Takeaways

  • Usage Restrictions: Anthropic is moving to limit how Claude Code interacts with third-party harnesses.
  • Subscription Misalignment: Current subscription plans were not built to support the high-intensity or specific usage patterns of external tools.
  • Official Confirmation: The news was confirmed by Boris Cherny, the head of Claude Code, through social media.

In-Depth Analysis

The Rationale Behind Usage Limits

Boris Cherny, the head of Claude Code, has clarified the reasoning behind the upcoming restrictions on third-party tool integration. The core issue lies in the architecture of Anthropic's subscription models. According to Cherny, these tiers were developed with specific user behaviors in mind, which do not align with the automated or high-frequency usage patterns often seen when Claude Code is utilized through third-party harnesses. By restricting these integrations, Anthropic appears to be protecting the integrity of its service delivery and ensuring that the resource consumption remains within the bounds of its designed business model.

Impact on Third-Party Harnesses

Third-party harnesses, which often wrap AI models into specialized developer environments or automation workflows, represent a significant portion of the advanced developer ecosystem. However, because these tools can trigger usage spikes that exceed the expectations of standard subscription plans, Anthropic has identified a need to decouple Claude Code from these external environments. This decision suggests that the current subscription framework lacks the flexibility to handle the "harness" style of interaction without potentially compromising service stability or financial sustainability for the provider.

Industry Impact

This move by Anthropic signals a growing trend among AI providers to exert more control over how their models are consumed via external platforms. As the industry matures, the gap between "direct-to-consumer" subscriptions and "API-like" usage through third-party tools is becoming a point of friction. For the AI industry, this could lead to more specialized subscription tiers specifically designed for automated harnesses, or it may force third-party developers to seek deeper, more formal partnerships with model providers to ensure continued access for their user bases.

Frequently Asked Questions

Question: Why is Anthropic restricting Claude Code on third-party tools?

According to Boris Cherny, the head of Claude Code, the current subscriptions were not designed to handle the specific usage patterns associated with third-party harnesses.

Question: Who announced these changes?

The announcement was made by Boris Cherny, the head of Claude Code at Anthropic, via the social media platform X.

Read Original Article

Related News

Meituan LongCat Releases General 365: A Challenging New Benchmark for AI Reasoning Evaluation
Industry News

Meituan LongCat Releases General 365: A Challenging New Benchmark for AI Reasoning Evaluation

Meituan's LongCat team has officially open-sourced General 365, a new evaluation benchmark designed to measure the reasoning capabilities of large language models (LLMs). In a comprehensive test involving 26 mainstream models, the results revealed a significant gap in current AI reasoning performance. Even the top-performing model, Gemini 3 Pro, achieved an accuracy of only 62.8%, while the vast majority of tested models failed to reach the 60% passing mark. This release aims to establish a more rigorous standard for the industry, highlighting the current limitations of even the most advanced AI systems in complex reasoning tasks. By providing a transparent and difficult metric, Meituan seeks to drive the development of more logically capable artificial intelligence.

Managing AI Coding with Agent Evaluation Thinking: Meituan's Practice in Refactoring 310,000 Lines of Code
Industry News

Managing AI Coding with Agent Evaluation Thinking: Meituan's Practice in Refactoring 310,000 Lines of Code

As AI-generated code now accounts for over 90% of development in certain environments, the primary challenge has shifted from generation speed to the effective management and constraint of AI capabilities. Meituan's technical team recently shared their experience refactoring 310,000 lines of code using a strategy centered on "Agent evaluation thinking." By implementing technical debt assessment, standardized rules, a specialized Refactoring SOP, and a Pre-PR (Pull Request) mechanism, they have successfully transformed large-scale refactoring from a high-cost, periodic project into a continuous, daily operational task. This approach ensures that AI-driven development does not amplify systemic chaos but instead adheres to unified technical standards, maintaining long-term code quality and system stability in an AI-dominated coding era.

Meituan Technical Team Releases LARYBench: A New Benchmark for Universal Latent Action Representation in Embodied AI
Industry News

Meituan Technical Team Releases LARYBench: A New Benchmark for Universal Latent Action Representation in Embodied AI

The Meituan Technical Team has officially introduced LARYBench (Latent Action Representation Yielding Benchmark), a systematic evaluation framework designed to guide the learning of universal latent action representations from large-scale visual data. This benchmark marks a significant milestone in embodied AI by providing a standardized way to measure how models learn actions from visual inputs. Experimental results from the benchmark reveal that general vision models significantly outperform specialized embodied action expert models in both action generalization and control precision. Furthermore, the research demonstrates that embodied action representations can naturally emerge from large-scale human video data, suggesting that broad visual training is a viable path toward achieving more sophisticated and adaptable robotic control systems.