I hate Opus 5. It’s the best model, anyway.

2026-07-24 · Source: How I AI · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, extended

Summary

The "How I AI" benchmark review of Anthropic's new Opus 5 model reveals a complex user experience despite top-tier performance. The author, initially frustrated by Opus 5's "neurotic" and timid personality, excessive verbosity ("Claude's slop"), and perceived human reliance, conducted a comparative "personality interview" with Opus 5 and GPT models. While Opus 5 articulated itself as a tool for volume and breadth, GPT emphasized speed and scale. Despite the interaction challenges, Opus 5 achieved the highest score in the "How I AI" benchmark, which evaluates tasks like PRD creation, prototype creation, and agentic coding. The model particularly excelled in front-end design and app prototyping, producing detailed, functional, and polished outputs, leading the author to conclude that its high-quality results outweigh the exasperating direct interaction.

Key takeaway

For Machine Learning Engineers or AI Scientists evaluating frontier models for design and prototyping, Opus 5, despite its verbose and timid interaction style, delivers superior output quality. You should consider integrating Opus 5 into your workflow for front-end design, app design, and prototype creation, especially for asynchronous tasks where direct conversational interaction is minimized. Prioritize evaluating models based on final output quality over conversational "personality" for specific use cases.

Key insights

Opus 5 excels in output quality for design tasks despite a "neurotic" and verbose interaction style.

Principles

• Model "personality" impacts user experience.
• Output quality can outweigh interaction friction.
• Benchmarks should include qualitative "vibe checks."

Method

The "How I AI" benchmark evaluates models on PRD, prototype, wireframe, bug triage, and agentic coding, using a 70% human "vibe score" and 30% LLM-as-judge split.

In practice

• Use Opus 5 for front-end design and prototyping.
• Compare model "personalities" for specific workflows.
• Integrate human judgment with LLM-as-judge evaluations.

Topics

• Opus 5
• LLM Evaluation
• AI Benchmarking
• Front-end Prototyping
• Model Personality
• GPT Comparison

Best for: AI Architect, AI Engineer, CTO, AI Scientist, Machine Learning Engineer, Director of AI/ML

Related on AIssential

• Opus 5 is my new go-to model — Opus 5 balances Fable's quality and Soul's diligence, offering a safer, more cost-effective, and instruction-following model for daily use.
• Claude Opus 4.8 is here. Is it as good as they say? — Opus 4.8 excels in one-shot coding but struggles with complex edge cases, existing codebases, and factual grounding.
• 🎙️ How I AI: Claude Opus 5 Review + Browser use in Codex + How Cursor and a Raspberry Pi makes AI fun — AI's utility extends to browser automation, hardware development, and nuanced model personality assessment, revealing new operational efficiencies.
• Opus 5 — Claude Opus 5's practical coding and agentic capabilities appear to exceed its initial aggregate benchmark scores, highlighting evaluation challenges.
• GPT 5.5 vs Opus 4.7: Which is the Best AI Model Today? — GPT-5.5 excels in agentic tasks and direct communication, while Claude Opus 4.7 leads in creative writing and coding.
• OPUS 5 CLICK NOW — Claude Opus 5 redefines LLM value, offering superior performance at a significantly lower cost per task.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by How I AI.