Anthropic Launches Claude Opus 5 as Cost-Efficient Alternative to Fable 5

Anthropic has released Claude Opus 5, a model that matches or exceeds its flagship Fable 5 on most coding and knowledge-work benchmarks at half the token price Source: The Next Web. Priced at $5 per million input tokens and $25 per million output tokens, Opus 5 is designed for everyday use, offering near-frontier intelligence without the cybersecurity restrictions that limit Fable 5 Source: Ars Technica.

Performance and Benchmarks

• On Frontier-Bench, Opus 5 more than doubles its predecessor's score and surpasses every competing model, including Fable 5 Source: The Next Web.
• On ARC-AGI 3, it scores roughly three times higher than the next-best model Source: The Next Web.
• On Zapier AutomationBench, it passes more tasks than any rival even at its lowest effort setting Source: The Next Web.

Ars Technica notes that while Opus 5 shows iterative improvements, it is not a radical capability leap, but rather focuses on token efficiency Source: Ars Technica.

Safety and Alignment

Anthropic claims Opus 5 is its safest model yet, with the lowest rates of deceptive behavior and susceptibility to misuse Source: Gizmodo. However, due to intentional training decisions, it lags behind Mythos 5 on offensive cybersecurity tasks, though it has improved at finding vulnerabilities as a side effect of general capability gains .