Anthropic, a San Francisco-based artificial intelligence company, unveiled Claude Opus 5 on Friday, positioning the new model as a cost-effective alternative to its most advanced AI systems. The company claims Opus 5 delivers performance close to its highest-tier model, Fable 5, while reducing usage costs by approximately 50 percent.
The release comes amid growing concerns among business clients over the high expenses associated with deploying cutting-edge AI tools. Anthropic said Claude Opus 5 outperforms industry standards in coding and knowledge-based tasks while requiring fewer computational resources compared to previous versions.
In pricing terms, Opus 5 users will pay $5 per million input tokens and $25 per million output tokens, half the rate of the existing Fable 5 model. Anthropic’s pricing is based on token usage, a measure common in the AI sector where customers are charged according to the volume of data processed through the system.
Claude Opus 5 directly competes with other AI models such as OpenAI’s ChatGPT and Google’s Gemini, both of which also use token-based billing structures. The timing of the launch coincides with intensified industry scrutiny following recent stock market filings by Anthropic and OpenAI, placing the companies’ financial models and growth prospects under closer examination by investors.
On security and safety issues, Anthropic disclosed that Opus 5’s training deliberately excluded cyber offensive capabilities. The firm differentiates it from Mythos 5, the most powerful and restricted AI in its portfolio, which remains accessible only to select partners. Mythos 5 is designed with advanced capabilities for identifying vulnerabilities in computer systems and has sparked controversy due to its potential use in developing cyberattacks.
While Anthropic stated that Opus 5 matches Mythos 5’s proficiency in detecting software flaws, it lags significantly in converting those vulnerabilities into actionable threats. This reduced risk profile is reflected in the company’s assertion that automated safety filters intervene on Opus 5 usage approximately 85 percent less frequently than with Fable 5, which belongs to the same class as Mythos.
In June, both Fable 5 and Mythos 5 were temporarily removed from public access following export control restrictions imposed by the U.S. Commerce Department. These controls were lifted later that month, allowing the models’ availability to resume.
The introduction of Claude Opus 5 marks a significant step for Anthropic as it seeks to balance performance, cost efficiency, and security in an increasingly competitive and regulated AI market.