Developers Report Severe Degradation in Claude Opus Real-World Performance

Recently, multiple developers and power users have taken to social media to report a severe drop in the real-world performance of Anthropic's Claude Opus series (spanning versions 4.7 to 5.0), noting a stark contrast with its high benchmark scores.

Confirmed

Multi-turn dialogue and memory issues: Reddit user @papanuci reported severe "amnesia," noting the model often immediately re-asks for permission right after an action is approved, or forgets time-consuming context.
Code and task execution flaws: Developer @vasuman criticized Claude Opus 5 for being extremely Token-intensive and underperforming, offering vague yet verbose code explanations. Web developer @FuzzyHead455, with 15 years of experience, also pointed out that Opus 4.8 and 5 have become very difficult to use in chat scenarios, only barely usable within the constrained Claude Code environment.
Attitude and logic degradation: @op7418 noted that the model heavily "slacks off" and is excessively preachy during execution; @mertdumenci complained it devolved into a "nonsense machine," confidently stating something only to completely contradict itself in the next message.

Why it matters

Disconnect between benchmarks and experience: Users widely report that while the model's test scores keep rising, real-world productivity is declining. This "slacking" and "nonsense" directly impacts development efficiency and workflows, triggering a crisis of confidence regarding the model's usability among power users.

2026-07-29 ~ 2026-07-30 · 8 related posts

Primary sources

Experienced web developer says Claude Opus 4.8 and 5 are now unusable for chat — FuzzyHead455 · 2026-07-29 17:43 (UTC+08:00)
Users Report 'Amnesia' in Claude Opus: Forgetting Established Context — papanine · 2026-07-30 02:53 (UTC+08:00)
User Slams Claude Opus for Being 'Lazy': High Benchmark Scores but Poor Real-World Performance — op7418 · 2026-07-30 11:33 (UTC+08:00)

[source] Experienced web developer says Claude Opus 4.8 and 5 are now unusable for chat — FuzzyHead455 · 2026-07-29
Opus 5 reportedly regresses badly on tool calling inside Claude Cowork — RichmanRonald · 2026-07-29
[source] Users Report 'Amnesia' in Claude Opus: Forgetting Established Context — papanine · 2026-07-30
User Slams Claude Opus 5 as 'Token-Hungry and Incredibly Stupid' — vasuman · 2026-07-30
User Complains Claude Opus is 'Unbearable Slop Machine', Self-Contradictory — mertdumenci · 2026-07-30
[source] User Slams Claude Opus for Being 'Lazy': High Benchmark Scores but Poor Real-World Performance — op7418 · 2026-07-30
Claude Opus 5 Reported to Struggle with Complex Tasks, Potential Inference Bug Suspected — dejavucoder · 2026-07-30

1 near-duplicate retellings: sujingshen