Opus 5: No-Hype Full Review & Testing

Value of the Information & Strength of the Argument

The video provides substantial value by moving beyond official benchmarks and offering practical, real-world testing across diverse creative and knowledge-based tasks. The argumentation is solid: the creator uses a consistent methodology (identical prompts, blind ranking, cost tracking) and supports his claims with visual evidence and detailed cost data. He also acknowledges potential biases and limitations, such as the influence of prompt phrasing and the small number of tests, which strengthens the credibility of his conclusions. The comparison is thorough, and the conclusion that Opus 5 outperforms Fable 5 is well-supported by the presented evidence.

Scientific Rigor, Source Quality, Title Accuracy

The video demonstrates a rigorous approach to testing, with a clear methodology and transparent reporting of results. The creator references Anthropic’s official benchmarks and provides links to his own blog post with prompts and generations, which adds to the transparency. However, the evaluation is based on a limited number of tests and is inherently subjective, as the ranking relies on the creator’s personal judgment. The title accurately reflects the content, and the video does not overhype or make unsupported claims. The creator also discusses the cost implications, which is a valuable addition. Overall, the sources are credible (official benchmarks, personal testing), and the title is appropriate.

The title accurately reflects the content: a comprehensive, no-hype review and testing of Opus 5.

8/10

The video presents a structured, hands-on evaluation of Opus 5 against Fable 5 and Opus 4.8, with transparent methodology (blind tests, multiple scenarios) and detailed cost analysis. The creator acknowledges limitations (e.g., small sample, potential bias in prompts) and does not overstate conclusions. However, the evaluation is subjective and not peer-reviewed, and the creator's own benchmarks are not independently verified.

• 00:00Intro
• 00:51Benchmarks
• 04:21Test 1: Web design
• 10:48Test 2: 3D and simulation
• 14:57Test 3: Playable game
• 17:44Test 4: Motion graphics
• 21:07Test 5: Knowledge work
• 27:10Verdict

• AI Bootcamp — Mentioned in the description as a four-week live bootcamp built on your company.
• AI for Mortals Newsletter — Linked in the description for subscribing to the newsletter.
• Opus 5 vs Opus 4.8 vs Fable 5 - Prompts and Generations — Blog post containing all prompts and generations used in the video.

• Anthropic's official benchmarks for Opus 5 — The video references Anthropic's official benchmarks, which claim Opus 5 outperforms Fable 5 on many metrics.

• Commenter claims Mythos 5 and Fable 5 are the same model — A commenter suggests that Mythos 5 and Fable 5 are the same underlying model, which contradicts the video's implication that they are distinct.