Google AI Studio recently launched the Gemini API's "Real-Time Connection to Google Search" feature, a move aimed at improving the response accuracy of the Gemini 1.5 model, reducing false information, and providing developers with a more reliable AI model. This feature allows Gemini models to access real-time data searched by Google, providing more accurate, transparent and up-to-date information. Developers can access this feature through Google AI Studio or Gemini API and flexibly control real-time data retrieval as needed to optimize the answer quality of the model. The launch of this feature also marks that Google is actively improving the competitiveness of its AI models in competition with competitors such as OpenAI.
With this update, the Gemini 1.5 model can get the latest information from Google searches, thereby enhancing the accuracy and transparency of answers. Developers can access this feature directly through the Tools section of Google AI Studios, or enable the "google_search_retrieval" tool in the Gemini API. The “Real-Time Connection to Google Search” feature is currently available in the free trial of Google AI Studios, while the Gemini API is priced at $35 per 1,000 queries with real-time connections. The main advantage of this feature is the ability to reduce false information in model answers and provide users with relevant and up-to-date information. In addition, the feature supports linking answers with real-time search results, improving transparency and providing traffic-oriented information sources.
Developers can activate active search functions for use when specific queries require real-time data. This feature predicts when real-time data is accessed based on the score, which will improve the response quality, allowing developers to control the use of real-time connections based on the set threshold.
In Google AI Studio's comparison mode, developers can see how real-time connections improve the quality of answers. Answers without real-time connections will only reflect the model knowledge deadline, while answers with real-time data will provide the latest information, support links, and search suggestions.
The launch of Google's feature comes as OpenAI released its SearchGPT. As ChatGPT, the latter enhanced its web search capabilities and provided users with timely and accurate answers, combining natural language interactions with the latest data, including sports, news, and stock markets. Information in areas such as quotations.
In addition, Google Gemini's real-time connectivity feature will compete with platforms such as Perplexity AI and ChatGPT's web search. Recently, the head of Perplexity AI also announced that the platform will soon launch new updates. Meanwhile, Meta is developing its own search engine, planning to integrate it into social media platforms such as WhatsApp and Instagram to provide real-time information services.
Key points:
The new feature "Real-time Connection to Google Search" is designed to improve the response accuracy of AI models.
Gemini API is priced at $35 per 1,000 queries, and supports real-time data retrieval.
Developers can flexibly activate real-time data retrieval according to their needs to improve the quality of answers.
All in all, Google's move aims to enhance the competitiveness of the Gemini model, gain an advantage in competition with other AI platforms, and provide developers with more powerful tools. The launch of this feature will further promote the advancement of AI models in information acquisition and processing.