Anthropic's Claude Opus 5 landed on 24 July 2026, and it matters for any firm that wants to be named when a client asks an assistant for a recommendation.
Anthropic released a step-change model on 24 July
On 24 July 2026, Anthropic announced Claude Opus 5, calling it "a thoughtful and proactive model" and a clear step up from the previous Opus 4.8. The release also appears on Anthropic's running news feed, dated the same day. The company points the model at long-running agents, coding, knowledge work and scientific research.
The claims are specific. On Frontier-Bench, a software-engineering test, Anthropic says the new model more than doubles Opus 4.8's performance at a lower cost per task. On ARC-AGI 3, a benchmark built around problems a model has not seen before, it scores three times as high as the next-best model. Anthropic also reports gains on business automation and computer use, and says the model is much better at checking its own work and iterating until it succeeds. Visual output is stronger too.
The price holds steady. Claude Opus 5 costs the same as Opus 4.8: $5 per million input tokens and $25 per million output tokens, with a Fast mode at double the base price. It is available now on Claude.ai, Claude Code, Claude Cowork and the Claude API, under the model ID claude-opus-5.
Why a model release belongs on your radar
Your clients and prospects increasingly ask an assistant before they ask you. When someone asks Claude or a similar tool to suggest a firm for a particular job, the answer is shaped by what these models can read, trust and cite from the open web. A more capable, more widely used assistant means more of those first conversations happen before a person ever reaches your site.
You do not need to do anything with Claude Opus 5 itself. What matters is whether your firm is easy to find, clearly described and well cited online, because that is the raw material an assistant draws on when it answers. These models keep getting better at reading and reasoning over what is published. The firms that surface are the ones whose expertise sits somewhere an assistant can see it.
None of this is a fire drill. It is one more step in a direction that has been clear for a while, and a reminder that being genuinely findable is the move that lasts.
Frequently asked questions
Do I need to switch to or buy Claude Opus 5?
No. This is an update to an AI assistant, not a product your firm has to adopt. It is worth knowing because these assistants increasingly shape how buyers find and shortlist firms.
Will this change whether my firm gets found?
Not directly, and not overnight. A stronger model does not rank you higher on its own. What it reinforces is the value of being clearly published and well cited on the web, since that is what assistants like Claude read when they answer a buyer's question.
What actually changed on 24 July 2026?
Anthropic released Claude Opus 5, a more capable model than Opus 4.8, aimed at coding, knowledge work, research and long-running agents. It is available on Claude.ai, Claude Code, Claude Cowork and the Claude API.
Is it more expensive than the previous version?
No. Anthropic set the same pricing as Opus 4.8: $5 per million input tokens and $25 per million output tokens. A Fast mode is available at double the base price.
