Powered by Smartsupp

Anthropic Launches Opus 5: Outperforms Fable 5 on Key Benchmarks, Cheaper & Less Restrictive



By admin | Jul 24, 2026 | 2 min read


Anthropic Launches Opus 5: Outperforms Fable 5 on Key Benchmarks, Cheaper & Less Restrictive

On Friday, Anthropic unveiled its Opus 5 model, the latest iteration of its flagship heavyweight offering. Although it's less powerful than the Fable 5 model, Opus 5 is more affordable and comes with fewer restrictions, making it the preferred choice for most applications. Notably, Opus 5 surpasses Fable 5 on several benchmarks highlighted in the announcement. This launch comes just two months after Opus 4.8, which became available on May 28. The Mythos 5, Fable 5, and Sonnet 5 models all debuted in June, leaving only the lightweight Haiku model awaiting an upgrade to the 5 series.

In a blog post introducing the new model, Anthropic emphasized that Opus 5 is "much stronger at verifying its work and iterating carefully until it succeeds." They cited benchmark testing where Opus 5 independently developed a computer vision pipeline in response to an incomplete prompt, among other examples. Crucially, Opus 5 also avoids many of the restrictions that have plagued Fable since its release. Like its predecessor, Opus 5 is not subject to the 30-day data retention policy that applies to Fable and Mythos, which had raised privacy concerns among some users.

However, Opus 5 still includes meaningful safeguards, particularly around cybersecurity tasks such as exploit generation and penetration testing. For example, Opus 5's safeguards prevent it from being used to scan for vulnerabilities in a software binary, though it is allowed to search for vulnerabilities in source code, as the latter is more likely to serve defensive purposes. Overall, Anthropic expects these classifiers to engage 85% less frequently for Opus 5 compared to Fable 5, reflecting a lighter touch for the less capable model.

EMBED_PLACEHOLDER_0

Anthropic is also introducing a new tool to reduce disruptions when safeguards do activate. Users can now opt into a beta feature called Automatic Fallbacks, which automatically routes requests to a less powerful model when a prompt triggers the safety classifier. As a result, API users who enable this setting will receive a functional response instead of an error message.




RELATED AI TOOLS CATEGORIES AND TAGS

Comments

Please log in to leave a comment.

No comments yet. Be the first to comment!