Search Grounding
A technique in which a model retrieves information from external knowledge sources—such as web pages or image databases—in real time during content generation, using the retrieved results to constrain its output. The approach is used to reduce hallucinations and improve the factual accuracy of generated content.
Timeline (last 90 days)
Search grounding means the model actively calls a search engine to obtain real-time information when generating a response and attaches the results to the answer as supporting evidence, usually presented in the form of citation links.
Nano Banana 2 has native Web/image search grounding capabilities, able to reduce hallucinations when generating factual infographics or landmark structures by leveraging search results.
Search grounding refers to a model retrieving external knowledge sources in real time during content generation and using the retrieved results as contextual constraints to reduce hallucinations.
Retrieval methods have high accuracy but incur selection costs and context costs per query, and those costs balloon as user history grows.
All Facts (4)
Search grounding means the model actively calls a search engine to obtain real-time information when generating a response and attaches the results to the answer as supporting evidence, usually presented in the form of citation links.
50%UnverifiedNano Banana 2 has native Web/image search grounding capabilities, able to reduce hallucinations when generating factual infographics or landmark structures by leveraging search results.
50%UnverifiedSearch grounding refers to a model retrieving external knowledge sources in real time during content generation and using the retrieved results as contextual constraints to reduce hallucinations.
50%UnverifiedRetrieval methods have high accuracy but incur selection costs and context costs per query, and those costs balloon as user history grows.
50%