Anthropic launches Claude web search API, betting on the future of post-Google information access

Anthropor has provided an online search capacity to help her in Claud AI, which raises competition in the advanced AI’s advanced search market as technology giants are racing to redefine how to find information online.

The company announced today that developers can now enable Claude to access current web information through its application programming interface, allowing the artificial intelligence assistant to perform multiple progressive searches to collect complete comprehensive answers with source censorship. The move comes at a time when searching on the Internet is under its most important transformation because Google has made a revolution in this field for more than two decades.

Anthropor said in its announcement: “The developers can now increase the knowledge of Claude comprehensive about the current real data by enabling the web search tool when submitting requests to the messaging applications interface.”

The new capacity reaches amid signs that traditional research loses the land to the alternatives operating in Amnesty International. Today, the Vice president of Apple Services, Eddy Cue, witnessed today at the Google Anti -monopoly trial that retreated in Safari last month for the first time in the history of the 22 -year -old browser. “I have lost a lot of thinking about it,” Koy said regarding the loss of possible revenues from the estimated Google batch of $ 20 billion to be the SAFARI virtual search engine.

Search on the Internet is now available on our application programming interface.

The developers can increase the Klude’s comprehensive knowledge of modern data. pic.twitter.com/prqf0zkxuz

Anthropicai May 7, 2025

Artificial Intelligence Assistants eat Google lunch: declining the dominance of the traditional search engine

Data indicates a seismic shift in information discovery patterns. The Soci Consumer Index explains that 19 % of consumers are already using the artificial intelligence of the research, creating the first quiet calm challenge of Google on the arrival of web information decades ago.

This shift stems from the basic differences in how to process artificial intelligence assistants and provide information. Unlike traditional search engines that offer a list of links that require users to search for results, AI assistors collect information from multiple sources, and provide brief, contextual answers. This removes the cognitive load to evaluate many websites and extracts the exact information that users are searching for.

The timing of a human advertisement is especially important. With Safari’s searches for the first time ever – an unprecedented metal braid in his testimony – we are witnessing early indicators of the conversion of collective consumer behavior. Traditional traditional search engines are increasingly exceeding advertising revenue in favor of conversation -based interactions that give priority to the quality of information on commercial interests.

Under the cover: How API converts from Anthropor to retrieve information to developers

The artistic approach to anthropologist is a great progress in how to spread artificial intelligence systems as information gathering tools. The system uses a sophisticated layer to make decisions that determine when external information will improve response quality, generating targeted search queries instead of just passing the user’s questions to the back interface of the search.

This “Agentic” ability – allowing Claud to perform multiple progressive searches using previous results to inform subsequent queries – a more comprehensive search process than traditional search. Implementation mainly mimics how the human researcher can explore a topic, starting with general quarrels and gradually improving it based on the initial results.

For developers, the application programming interface provides likable control through max_uses A teacher who limits the number of sequential searches that Claude can perform. This treats both cost considerations and prevents artificial intelligence from falling into the rabbit’s research holes. The field control features provide decisive degrees for the deployment of institutions, allowing institutions to ensure that information comes only from reliable sources.

At 10 dollars per 1000 search operations in addition to standard symbolic costs, Antarbur put this feature as a distinct offer, indicating confidence in the value proposal justifies the price point compared to direct integration with free search engines.

AI Search Wars Heat Up: Big Tech Scrambles to redefine access to digital information

The competitive scene on artificial intelligence research has become increasingly crowded and controversial. Openai Integrated Web Search in Chatgpt last fall and expanded it recently with shopping features, and benefited from the base of huge users – 800 million weekly users, according to Openei SAM Altman – to challenge the dominance of commercial search from Google.

Apple’s potential axis is the most important threat to the current situation. With the Safari account by 17.25 % of the global browser use and 22.32 % on mobile devices according to the latest data from Statcountter, any step to replace Google with artificial intelligence search alternatives will significantly change the dynamics of the market. Apple’s discussions with Ai Perplexity, Openai and Anthropic indicate that the company is actively exploring multiple partnerships instead of developing its research technology.

Meanwhile, specialized players such as Perplexity AI make strategic moves to include themselves directly in the ecosystem of devices. Its partnership with Motorola is an early attempt to set AI search as a feature at the device level instead of just an application or website. Google’s reported resistance to this partnership, which was revealed during anti -monopoly procedures, indicates that the research giant recognizes the existential threat posed by these new models.

The dismantling of the court may be on the Safari deal from Google-which may eliminate up to $ 20 billion in Apple’s annual payments-speeding up this transition by removing the Apple’s financial incentive to maintain the current situation.

Content economy is disrupted: Publishers face an existential threat, as artificial intelligence exceeds traditional sites

The shift to artificial intelligence research represents deep challenges to the content economy that has evolved around traditional search engines. When AI assistants provide direct answers from multiple sources, they greatly reduce clicking to original content sites. This threatens the advertising business model that maintains a lot of the online information system.

The Vice President to search Google Pandu Nayak highlights the inability of “any guarantees” about the recovery of traffic on the attractiveness of this situation. Future content creators face artificial intelligence that may become more important than appearing in the search results – however, the custody alone does not generate advertising revenues that maintain content creation.

This creates an unsustainable dynamic: AI relies on high -quality web content to create responses, but by redirecting the user’s attention away from source web sites, they undermine the economic model that produces this content. Without a new compensation mechanism for content creators, long -term health of the ecosystem of information can be penetrated.

For companies, this transformation requires the basic rethinking of the digital strategy. The improved content of traditional search engines may perform badly in the discovery of artificial intelligence, while the content that focuses on providing clear and reliable information may gain prominence even without the signs of traditional economic officials.

Anthropic’s Web Search Application interface is more than just another feature in the artificial intelligence set-which indicates the development of access to Internet information towards a more integrated model based on conversation. When the search capabilities become included in artificial intelligence assistants through applications, a distinction continues to be a distinction, browsing and asking questions.

This rapprochement is evident in how to merge Quora to search the web into the Poe platform and how Adaptive.ai plans to use it in applications facing customers. Instead of research as a separate activity, the surrounding capacity is woven throughout the digital experience.

For ordinary users, this can eventually eliminate the need to think about the location of information search – either on Google, in a specialized application or through AI’s assistant. The cognitive load of the movement of internet information from humans will turn into artificial intelligence systems.

The race developed to redefine the research into a much larger thing: a competition to create the primary interface through which people reach digital information. As the traditional search model from Google stumbles, Apple is likely to prevent its long -term partnership of Google, and companies such as Anthropor open their capabilities to the ecosystem of developers, we are witnessing the most important restructuring to access information since the emergence of modern search engines.

In this new model, the winner will not be determined by those who have the best research algorithm, but by those who create the most trusted artificial intelligence interface capable of world information. It may soon be remembered to mention the era of keyword writing in a search box as just a transitional stage in our relationship with digital knowledge.