google/tools

12 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Google Code Execution tool

Google models can generate and execute Python code via the Code Execution tool. Enable it by adding the `code_execution` tool: `const { text, toolCalls, toolResults } = await generateText({ model: google('gemini-2.5-pro'), tools: { code_execution: google.tools.codeExecution({}) }, prompt: 'Use python to calculate the 20th fibonacci number.' });`

Google Search grounding tool

Google models have access to latest information using Google Search grounding via the `google_search` tool. Example: `const { text, sources, providerMetadata } = await generateText({ model: google('gemini-2.5-flash'), tools: { google_search: google.tools.googleSearch({}) }, prompt: 'List the top 5 San Francisco news from the past week. You must include the date of each article.' });`

Google Search tool configuration options

The `googleSearch` tool accepts optional configuration: searchTypes object (with webSearch and/or imageSearch, pass {} to enable; webSearch is default) and timeRangeFilter object (requires startTime and endTime in ISO 8601 format).

Google Search tool grounding metadata fields

When Google Search grounding is enabled, grounding metadata includes: webSearchQueries (string[] | null), searchEntryPoint (object with renderedContent | null), groundingSupports (array with segment, groundingChunkIndices, confidenceScores | null).

Google File Search tool

The File Search tool lets Gemini retrieve context from indexed documents in File Search stores. Example: `const { text, sources } = await generateText({ model: google('gemini-2.5-pro'), tools: { file_search: google.tools.fileSearch({ fileSearchStoreNames: ['projects/my-project/locations/us/fileSearchStores/my-store'], metadataFilter: 'author = "Robert Graves"', topK: 8 }) }, prompt: 'Summarise the key themes of I, Claudius.' });`

Google URL Context tool

Google provides a URL context tool that allows providing specific URLs for the model to analyze directly. You can add up to 20 URLs per request. Supported for Gemini 2.0 Flash models and above. Example: `const { text, sources, providerMetadata } = await generateText({ model: google('gemini-2.5-flash'), prompt: 'Based on the document: https://ai.google.dev/gemini-api/docs/url-context. Answer this question: How many links we can consume in one request?', tools: { url_context: google.tools.urlContext({}) } });`

Google URL Context metadata fields

URL context metadata includes urlMetadata (array of objects with retrievedUrl and urlRetrievalStatus) and groundingMetadata. Possible urlRetrievalStatus values include URL_RETRIEVAL_STATUS_SUCCESS.

Google Maps grounding tool

Google Maps grounding gives models access to Google Maps data for location-aware responses. Example: `const { text, sources, providerMetadata } = await generateText({ model: google('gemini-2.5-flash'), tools: { google_maps: google.tools.googleMaps({}) }, providerOptions: { google: { retrievalConfig: { latLng: { latitude: 34.090199, longitude: -117.881081 } } } satisfies GoogleLanguageModelOptions }, prompt: 'What are the best Italian restaurants within a 15-minute walk from here?' });`

Google Maps grounding metadata structure

When Google Maps grounding is enabled, groundingMetadata includes groundingChunks with maps objects containing uri, title, and placeId fields.

Google Maps grounding supported models

Google Maps grounding is supported on Gemini 2.0 and newer models.

Google Interactions API built-in tools mapping

Provider-defined tools under `google.tools.*` map to Interactions tool descriptors: google.tools.googleSearch → google_search (web/image search grounding), google.tools.codeExecution → code_execution (Python execution), google.tools.urlContext → url_context (fetch URLs), google.tools.fileSearch → file_search (File Search stores), google.tools.googleMaps → google_maps (Maps grounding), google.computer_use → computer_use (browser environment), google.mcp_server → mcp_server (MCP passthrough), google.retrieval → retrieval (Vertex AI Search).

Google Interactions API function tools behavior

Function tools defined with the AI SDK `tool()` helper are translated to Interactions `function` tool descriptors. Other tool kinds emit a warning and are dropped.

Give your agent this brain