> ## Documentation Index > Fetch the complete documentation index at: https://aisa.one/docs/llms.txt > Use this file to discover all available pages before exploring further. # Anthropic Web Search > Ask a question and get a Claude-written answer grounded in a live web search, with the underlying sources and inline citations, billed at exact cost. Ask a question and get an answer that Claude wrote after searching the live web. Send a normal Messages request — a `messages` array plus `max_tokens` — and the endpoint injects a fixed, server-pinned model and the `web_search` tool for you; you cannot override the model or add tools, which keeps cost bounded. Claude decides when to search (up to `max_uses` searches, default 5), reads the results, and answers with inline citations. The response is a standard Anthropic Messages object: `content[]` contains `server_tool_use` (the queries issued), `web_search_tool_result` (the sources found) and `text` blocks (the answer with `citations`), and `usage.server_tool_use.web_search_requests` reports how many searches were billed. Billing is pay-as-you-go at exact cost: `web_search_requests × $0.01` plus the model's own token cost, with no markup; a failed search (HTTP 200 `web_search_tool_result_error`) is not billed. Use this when you want a written, cited answer grounded in current web content — for open-web research that returns ranked links and page text in one call use [`post_tavily_search`](/docs/api-reference/search/post_tavily-search) instead, and for the OpenAI-model equivalent see [`post_openai_websearch_search`](/docs/api-reference/search/post_openai-websearch-search). ## OpenAPI ````yaml openapi/websearch.json POST /anthropic-websearch/search openapi: 3.0.0 info: title: Web Search API version: 1.0.0 description: >- Model-grounded web search endpoints that run a real web search inside a fixed frontier model and return the model's answer together with the underlying search results and citations. Billed pay-as-you-go at exact provider cost (no markup). servers: - url: https://api.aisa.one/apis/v1 security: - bearerAuth: [] paths: /anthropic-websearch/search: post: tags: - >- https://docs.anthropic.com/en/docs/build-with-claude/tool-use/web-search-tool summary: Model-grounded web search (Anthropic). description: >- Ask a question and get an answer that Claude wrote after searching the live web. Send a normal Messages request — a `messages` array plus `max_tokens` — and the endpoint injects a fixed, server-pinned model and the `web_search` tool for you; you cannot override the model or add tools, which keeps cost bounded. Claude decides when to search (up to `max_uses` searches, default 5), reads the results, and answers with inline citations. The response is a standard Anthropic Messages object: `content[]` contains `server_tool_use` (the queries issued), `web_search_tool_result` (the sources found) and `text` blocks (the answer with `citations`), and `usage.server_tool_use.web_search_requests` reports how many searches were billed. Billing is pay-as-you-go at exact cost: `web_search_requests × $0.01` plus the model's own token cost, with no markup; a failed search (HTTP 200 `web_search_tool_result_error`) is not billed. Use this when you want a written, cited answer grounded in current web content — for open-web research that returns ranked links and page text in one call use [`post_tavily_search`](/api-reference/search/post_tavily-search) instead, and for the OpenAI-model equivalent see [`post_openai_websearch_search`](/api-reference/search/post_openai-websearch-search). operationId: post_anthropic_websearch_search requestBody: required: true content: application/json: schema: type: object required: - messages - max_tokens properties: messages: type: array description: >- Conversation messages, same format as the Anthropic Messages API. The user turn holds your question. items: type: object properties: role: type: string enum: - user - assistant content: type: string description: >- Message text. Structured content blocks are also accepted. example: - role: user content: >- What are the three biggest AI announcements this week? Give one sentence each with sources. max_tokens: type: integer description: >- Maximum number of tokens to generate in the answer. Required by the upstream Messages API. default: 1024 example: 1024 system: type: string description: Optional system prompt to steer the answer's tone or format. example: max_tokens: 1024 messages: - role: user content: >- What are the three biggest AI announcements this week? Give one sentence each with sources. responses: '200': description: >- A standard Anthropic Messages response. `content[]` interleaves the searches performed and the cited answer; `usage.server_tool_use.web_search_requests` is the billed search count. headers: X-AISA-Web-Search-Count: description: Number of web searches billed for this request. schema: type: integer X-AISA-Price-USD: description: Amount charged to your account for this request, in USD. schema: type: string content: application/json: schema: type: object properties: id: type: string type: type: string example: message role: type: string example: assistant stop_reason: type: string example: end_turn content: type: array description: >- Ordered blocks: server_tool_use (search queries), web_search_tool_result (sources), and text (the cited answer). items: type: object usage: type: object properties: input_tokens: type: integer output_tokens: type: integer cache_read_input_tokens: type: integer cache_creation_input_tokens: type: integer server_tool_use: type: object properties: web_search_requests: type: integer description: Billed web search count. components: securitySchemes: bearerAuth: type: http scheme: bearer ````