Artificial intelligence is advancing at a pace that’s hard to keep up with, and every announcement seems to promise a revolution. On July 24, 2026, Anthropic unveiled Claude Opus 5, its new flagship model, with a message that differs from the usual: it’s not so much about "more power" as about more efficiency. For a company evaluating AI projects, that nuance matters more than it seems. At Tisa, we read it this way: useful AI is no longer the one that’s most spectacular in a demo, but the one that performs every day at a reasonable cost. Let us explain what Claude Opus 5 means for your business.
What Claude Opus 5 is and why people are talking about "efficiency"
Claude Opus 5 is the new version of Anthropic’s most advanced model, one of the leading providers of generative AI alongside OpenAI, Google, and Microsoft. According to Anthropic’s official announcement, the model achieves top-tier results in programming and knowledge work tasks, and it’s designed to be "used every day" because it works more efficiently than its alternatives.
The key to the launch isn’t a price increase, but the opposite: it’s offered at the same cost as its predecessor (Opus 4.8), but with notably superior performance. Anthropic sets it at 5 dollars per million input tokens and 25 per million output tokens. In plain terms: you get more for the same.
More performance at the same price: what it means for an SME
Here’s the real news for private companies. For years, improving AI quality meant paying more. This launch points to an underlying trend: the relationship between intelligence and cost improves with each generation. For an SME that was waiting for "the numbers to add up," that break-even point is getting closer.
What tasks come into play? Drafting and reviewing documents, assisted customer service, data analysis, support for development teams, or automation of administrative processes. Cases that were previously expensive to sustain in production are starting to make economic sense.
The "effort" adjustment: paying only for the intelligence you need
A practical new feature: the model lets you regulate the "effort" it dedicates to each task. For complex work you can squeeze out its maximum capacity; for routine tasks you can lower the level to respond faster and spend fewer tokens. Translated into business terms, it’s like having a professional who dedicates more or less time depending on the difficulty of the assignment. Well configured, it avoids paying for surplus power where it isn’t needed: exactly the kind of decision that separates a profitable AI pilot from one whose cost spirals out of control.
From answering to executing: the era of agents
The other message of the launch is the maturity of AI agents: systems that don’t just answer questions, but execute multi-step tasks using tools. Anthropic highlights the model’s ability to verify its own work and iterate until it solves a problem, and cites examples in programming environments and in the automation of business processes.
For a company, this is more relevant than any laboratory result. A reliable agent can take charge of a complete workflow —receive a request, consult data, prepare a response, and leave the work ready for review— instead of merely making suggestions. It’s the leap from "a chat that helps" to "an assistant that works."
Reliability and security: what really matters in a project
In a business context, reliability weighs as much as power. According to Anthropic, Opus 5 is its "most aligned" model to date, with lower rates of deceptive behavior and greater resistance to misuse. The company also describes specific safeguards in sensitive areas such as cybersecurity.
How we ground it in your business
At Tisa we don’t sell "AI for the sake of having AI." We work on applied AI by measurable use case: we identify a specific problem, assess the available data, define governance, and develop a use case tailored to your business. Choosing the model —whether it’s Claude Opus 5, Microsoft’s Copilot family integrated into Business Central and Power Platform, or another— comes afterward, based on what you need to solve, not the other way around.
We are a Microsoft partner with over 30 years of experience, and that foundation allows us to connect AI with the ERP and the processes you already use every day. Because a brilliant model, isolated from your data and your processes, delivers little; value appears when it’s integrated into the company’s real workflow.
Conclusion: efficiency brings AI closer to your company
The launch of Claude Opus 5 confirms a clear direction: more capability, the same price, and models designed to actually work, not to impress in a demo. For private companies, it’s a good sign: adopting AI is increasingly affordable, as long as it’s done with judgment.
Want to know which AI use case would make sense in your company and how to ground it on your data and your ERP? Write to us at info@grupotisa.com or call us at (+34) 971 305 885 for a no-obligation assessment. At grupotisa.com we’ll tell you more.