Anyone else experiencing lower quality in Claude Code since a few weeks?

Question

Hi,I am on the mid-plan (paid) and using the web interface and in the last weeks the quality of Claude's output quality decreased dramatically. The model used is Sonnet 4.5 and my workflow is copy/pasting code and inspecting the diffs.It changes code it shouldn't change, even makes syntax errors (like a character different than the one provided in a source file).I know these LLMs change all the time, and this is just an anecdote, but I am interested what others think.

acqbu · Accepted Answer

No idea, I haven&rsquo;t used CC in a while but I&rsquo;ve noticed a general pattern with these AI tools where they seem to start strong then slowly degrade over time.Could be cost optimisation through some sort of resource throttling. Most of these companies aren&rsquo;t profitable so it wouldn&rsquo;t surprise me if they&rsquo;re quietly tuning things down to save compute.

mustaphah · Answer

Are you based in the US? I've heard that compute capacity peaks during US working hours, and access may be degraded during that time, for example, through dynamic quantization [1].[1] https://www.seangoedecke.com/ai-is-good-news-for-australian-...