[BUG] Claude Opus quality regression since ~January 2025

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

I'm a daily Claude Code user (Pro subscription) and have noticed a significant decline in Claude Opus output quality since approximately January 2025. Back then, Opus was clearly the best AI coding assistant available —
now it frequently produces subpar results that I can't rely on.

What Should Happen?

Claude Opus should return to the quality level it had around January 2025, where it was clearly the best AI coding assistant available. Responses should be reliably accurate, with strong reasoning about code architecture and fewer cases where I need to second-guess or redo its output.

Error Messages/Logs

Steps to Reproduce

1. Use claude Opus in any way as before - it doesn't matter

Claude Model

Opus

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

dosen't matter

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

• This is based on sustained daily usage across multiple large web development projects
• The regression is hard to quantify with specific before/after examples — it's a general decline in reasoning quality, code suggestions, and reliability
• I've started evaluating alternative tools (e.g. Codex) because the current quality doesn't justify the subscription cost

Considering cancelling my subscription if this doesn't improve in the near future. Opus was genuinely best-in-class in January and I'd like to see it return to that level.