Opus 5 is praised for idea diversity in research, but users say it feels more cautious
dejavucoder · x · 2026-07-26
Claude Opus 5 is being praised for idea diversity but feels more cautious
The reply says Opus 5 is “so good for introducing idea diversity,” and describes it as an excellent research peer. Compared with Opus 3, it is said to have lost some of the older model’s charm, but to make up for it with a more pragmatic and abrupt tone that works well for rigorous research.
The author also notes that it feels a little too cautious — as if it has been trained heavily on “Reviewer 2” — but that this caution can be useful if treated as a feature rather than a bug.
Related event: Claude Opus 5 Early Tests: Better Efficiency but Overly Proactive(29 posts)→
More from Models
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27
- ChatGPT's crawler sample code strips out all Schema markup — VeryWellVersed · 2026-10-27