We Tested Claude Opus 5. It’s Frustrating with Flashes of Brilliance.

By Every

Categories: Startup, Product, AI

Summary

Claude Opus 5 is argumentative and stops early on complex tasks, making it frustrating despite flashes of genius—but the real issue is that developers need to rewrite prompts and dial down reasoning levels instead of using high/max thinking for all tasks.

Key Takeaways

  1. Opus 5 breaks existing workflows designed for 4.8; it stops too early on autonomous loops and argues more frequently with complex skill files, requiring developers to rewrite prompts rather than upgrade.
  2. Model selection should be task-based: use GPT-4.o for 80% daily coding work, reserve Opus 5 for medium/low reasoning tasks, and keep a high-end model for 20% of complex autonomous projects.
  3. Reasoning level (low/medium/high/max) is now as critical as model family; using high reasoning on simple tasks triggers argumentative behavior, while medium reasoning often produces better results on Opus 5.
  4. Opus 5's personality mirrors Fable's opinionatedness without Fable's intelligence—it's only excusable if you're an expert; non-experts will find it more annoying than productive compared to 4.8.
  5. In 2026, AI tool optimization shifts from model selection to reasoning dial-tuning; the meta-skill is experimenting with thinking levels rather than just picking better models.

Related topics

Transcript Excerpt

It's model release day. Opus 5 is out. We have your day zero vibe check. So, let's get into it. We've been testing this model for about a week now and it's I got to say it's a little bit of a hard model to love. >> What have they done to my boy? >> Everyone was like like it's it's not that usable. It stops. It argues with you. But as we've been testing it, um there have been some glimmers of things where you're like, "Oh, there there might be something here." So, here's what we're going to talk about. We're going to talk about whether you should switch to it. if you're a codeex user, what you should expect if you're a cloud user, and we're going to talk about the state of the race. What does it mean for OpenAI and anthrop I feel like I have uh two big slats in my life right now for models.…

More from Every