@Mayhem4Markets: Hate to say it but I'm also not impressed with Opus 4.8. Feels weak compared to Opus 4.6 right after release. Before th…
Summary
A user criticizes Anthropic's Opus 4.8 model, comparing it unfavorably to Opus 4.6 and calling it weak; another user echoes the sentiment, calling the model 'stupid'.
View Cached Full Text
Cached at: 06/01/26, 01:03 AM
Hate to say it but I’m also not impressed with Opus 4.8.
Feels weak compared to Opus 4.6 right after release.
Before the inevitable brain drain.
What’s your experience been like, friends?
Ahmad (@TheAhmadOsman): Tried Opus 4.8 in different harnesses than Claude Code
and nah this model is just stupid, good job Anthropic lol
Similar Articles
Claude Fable 5 spotted on Azure and the backend, likely the public-facing version of Claude Mythos 5
Claude Fable 5, likely the public-facing version of Claude Mythos 5, has been spotted on Azure and the backend.
A Claude Code skill that sweeps your codebase for architecture-deepening opportunities (delta-aware, analysis-only) and files them as ready-to-grab issues
A Claude Code skill that autonomously scans codebases for architecture improvement opportunities (e.g., shallow modules), creates delta-aware, analysis-only issues and PRDs, ready for implementation.
Anthropic Says We Must Stop Authoritarian AI. But What About Its Authoritarian Investors?
Anthropic claims to oppose authoritarian AI but is partially owned by the repressive monarchy of Abu Dhabi, revealing a double standard in its ethical stance.
OpenAI files confidentially for IPO, following Anthropic
OpenAI has confidentially filed for an IPO, following rival Anthropic's similar move, signaling a major year for AI company public offerings. The filing comes amid high spending and missed revenue targets, with OpenAI valued at $852 billion.
@rohanpaul_ai: New Anthropic research shows AI agents may look brilliant at code, but in biology they can fail before the science star…
Anthropic research reveals that AI agents struggle with biology databases, producing highly variable answers for the same query (e.g., Ebola sequence counts ranging from 5 to 106 vs. expected 266), but adding a repeatable retrieval tool significantly improves consistency and accuracy.