genai / news / / VentureBeat
Developers and AI power users are taking to social media to accuse Anthropic of degrading the performance of Claude Opus 4.6 and Claude Code.
Anthropic changed Claude Opus 4.6's default reasoning effort to medium on March 3, 2026.
KEY POINTS
- A viral GitHub post by AMD's Senior Director analyzed 6,852 sessions, showing measurable performance regression since February.
- Anthropic confirmed a March 6 change reducing Claude Code's prompt-cache TTL from one hour to five minutes for some requests.
- Benchmark claims of Claude Opus 4.6's decline were challenged due to differing task counts and possible statistical noise.
- Anthropic denies secret model downgrades, attributing user complaints to disclosed product and interface changes.
COMPANIES
Summarized by Newsio from VentureBeat. How we summarize →