Claude Code Addresses Recent Quality Issues and Resets Usage Limits

Apr 23, 2026

AI Summary

Claude Code has resolved three identified issues that led to reports of degraded performance for some users. As of April 20, the service has been updated, and usage limits for all subscribers are being reset to improve user experience.

Over the past month, reports indicated that Claude Code's performance had worsened for some users, traced to three specific changes affecting Claude Code, the Claude Agent SDK, and Claude Cowork, while the API remained unaffected.
The issues were resolved by April 20, 2023, with a reset of usage limits for all subscribers initiated on April 23.
The first issue involved the default reasoning effort setting, which was initially set to high in the Opus 4.6 release, causing latency and token usage problems. Feedback led to a change to medium effort, which was later reverted back to xhigh effort for Opus 4.7.
The second issue stemmed from a bug in a feature meant to improve efficiency by clearing old reasoning history, which inadvertently caused Claude to forget context during sessions. This bug was fixed on April 10, 2023.
The third issue related to verbosity in the new Opus 4.7 model, which was addressed by reverting a system prompt change that negatively impacted performance.
To prevent future issues, the team plans to enhance internal testing processes, improve the Code Review tool, and implement stricter controls on system prompt changes.
The company expressed gratitude to users who provided feedback, which was crucial in identifying and resolving these problems.

claudecode qualityai modelsanthropicnlp

Claude Code Addresses Recent Quality Issues and Resets Usage Limits

Related Stories

Thinking Machines Lab develops AI model for simultaneous conversation

ChatGPT Sees Increased Adoption Among Older Users in Early 2026

Optimizing Matrix Multiplication for Swift in LLM Training

arXivLabs Encourages Collaboration on New Features with a Focus on Privacy