Anthropic unveiled Claude Opus 5 on Thursday, positioning the new model as a capable alternative to its more advanced Fable 5 system. The company stated that Opus 5 delivers performance approaching Fable 5 across numerous domains, with particular improvements in handling complex coding tasks. The release comes in the context of recent regulatory scrutiny following government concerns about the Fable 5 model and subsequent negotiations with the Trump administration over AI safety protocols.
The company indicated that Opus 5 underwent independent testing by government partners prior to public release. Anthropic is targeting enterprise customers with Opus 5, emphasizing its suitability for knowledge work, biology applications, and everyday use cases, while positioning Fable 5 as better suited for more ambitious projects and extended AI agent operations. The company emphasized that Opus 5 incorporates stronger cybersecurity safeguards compared to its predecessor, Opus 4.8, and described it as the most aligned model in the Opus lineup with reduced susceptibility to misuse.
Regarding pricing, Anthropic maintained the same per-token rates as Opus 4.8 at $5 for input and $25 for output tokens, positioning Opus 5 as half the price of Fable 5 and slightly less expensive than OpenAI’s GPT-5.6. The company introduced a “Fast mode” option launching in research preview that offers increased processing speeds at double the standard price. Additionally, users can enable automatic fallback mechanisms to lower-tier models when safeguards prevent Opus 5 from processing requests. These features appear designed to address ongoing industry discussions regarding usage caps and cost pressures facing AI customers.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI