I am looking to understand the costs involved in Gemini API when using Grounding with Google search feature
According to the documentation, the costs involved in grounding via Google search would include:
Once a Google search is performed with the input query, data will be fetched from relevant website links and used as context for the LLM. Are these context tokens considered as Input tokens as well? If so, can customers view the context provided or get an approximation of the tokens fetched as context?.
Additionally, what happens if the query is complex and requires multiple searches? Will this be treated as a single grounding request or multiple requests(based on the number of websearches), depending on the query? For example, consider the following query: “What were the main topics discussed in the recent G7 summit, and provide a brief write-up for each topic, along with the most recent news related to them.”
Hi @ManuJose,
Welcome to Google Cloud Community!
It appears that additional assistance is required for your project. For more detailed guidance, please contact Google Cloud Billing Support.
Was this helpful? If so, please accept this answer as “Solution”. If you need additional assistance, reply here within 2 business days and I’ll be happy to help.
User | Count |
---|---|
2 | |
1 | |
1 | |
1 | |
1 |