TLDR: Anthropic, the AI company behind the Claude Code assistant, has sparked widespread user frustration by implementing sudden and unclear usage limits, particularly affecting its premium subscribers. This unannounced change, coupled with broader network instability, has disrupted developer workflows and raised concerns about transparency and the reliability of AI services.
Anthropic, a prominent artificial intelligence company, is currently facing significant backlash from its user base following the unannounced implementation of stricter usage limits for its AI coding assistant, Claude Code. The changes, which reportedly began around Monday, July 14, 2025, have caused considerable disruption and frustration among developers, including those subscribed to the high-tier $200/month Max plan.
Users have taken to developer forums, such as GitHub, to voice their complaints, citing unexpected halts to their work and an inability to complete projects. Many expressed disbelief at how quickly their quotas were exhausted, with one paid user stating, “Your tracking of usage limits has changed and is no longer accurate. There is no way in the 30 minutes of a few requests I have hit the 900 messages.” Another Max plan subscriber noted, “In just a couple hours of work I’m getting ‘Approaching Opus usage limit…’ which is annoying. This started a couple days ago.”
The core of the user frustration stems from the lack of transparency surrounding these new limits. Anthropic’s official policy does not define absolute message numbers for its tiers, instead describing a relative system where paid plans offer multiples of a free tier whose own limit “will vary by demand.” This “floating-rate system” makes it nearly impossible for professional developers to forecast their access, turning mission-critical workflows into a gamble and creating significant business risk for those who rely on the service.
Compounding the problem, the stricter limits have emerged alongside broader network instability. Users have reported frequent API overload errors, receiving “Error 529” messages, despite Anthropic’s official status page sometimes indicating 100 percent uptime. When contacted by the press, an Anthropic representative offered a limited statement, acknowledging “slower response times” but failing to address the specific issue of usage limits or the lack of prior communication.
Also Read:
- GitHub Introduces Spark: AI Tool Revolutionizes Full-Stack App Development with Natural Language
- Anthropic Introduces AI Auditing Agents to Combat Model Misalignment
This situation highlights a growing tension between the immense operational costs of high-powered AI services and user expectations for reliable, predictable service, especially for premium subscribers. As the AI industry continues its rapid expansion, with rivals like OpenAI and Google constantly enhancing their models, the incident underscores the critical importance of clear and open communication in maintaining user trust and satisfaction.
