Web search in a chat - Writer AI Studio

Web Search Tool Overview

The web search tool for chat completions allows you to search the web for current information during a conversation. This guide explains how to use the web search tool in a chat completion and provides an example of how to use it.

You need an API key to access the Writer API. Get an API key by following the steps in the API quickstart. We recommend setting the API key as an environment variable in a .env file with the name WRITER_API_KEY.

Usage Example

This example uses the web search tool to find current information about AI developments during a chat completion. To use the web search tool:

  1. Create a tools array that specifies the web search tool.
    • The tool array should include the type and function parameters. The function parameter should include the include_domains and exclude_domains parameters.
  2. Create a messages array that contains the user message that prompts the model to use the web search tool.
  3. Call the chat.chat method with the tools parameter set to the tools array and tool_choice set to auto.

Example Code

cURL

curl --location 'https://api.writer.com/v1/chat' \
    --header 'Content-Type: application/json' \
    --header "Authorization: Bearer $WRITER_API_KEY" \
    --data '{\n        "model": "palmyra-x5",\n        "messages": [\
            {\n                "role": "user",\n                "content": "What are the latest developments in AI technology?"\n            }\n        ],\n        "tool_choice": "auto",\n        "tools": [\
            {\n                "type": "web_search",\n                "function": {\n                    "include_domains": ["wikipedia.org", "github.com", "techcrunch.com"],\n                    "exclude_domains": ["quora.com"]\n                }\n            }\n        ]\n    }'

Python

from writerai import Writer

client = Writer()  # Initialize the Writer client. If you don't pass the `api_key` parameter,
# the client looks for the `WRITER_API_KEY` environment variable.

messages = [{"role": "user", "content": "What are the latest developments in AI technology?"}]

response = client.chat.chat(
    model="palmyra-x5",
    messages=messages,
    tools=tools,  # The tools array defined earlier.
    tool_choice="auto"
)

print(response.choices[0].message.content)

JavaScript

import { Writer } from "writer-sdk";

const client = new Writer();  // Initialize the Writer client. If you don't pass the `apiKey` parameter,
// the client looks for the `WRITER_API_KEY` environment variable.

const messages = [{role: "user", content: "What are the latest developments in AI technology?"}];

const response = await client.chat.chat({
    model: "palmyra-x5",
    messages: messages,
    tools: tools, // The tools array defined earlier.
    tool_choice: "auto"
});

console.log(response.choices[0].message.content);

Understanding Tool Structure

The web search tool object has the following structure:

Parameter Type Description
type string The type of tool, which is web_search for the web search tool
function object An object containing the tool’s configuration
function.include_domains array An array of domains to include in the search results
function.exclude_domains array An array of domains to exclude from the search results

You can pass multiple custom tools in the same request. Prebuilt tools include the web search tool, knowledge graph tool, LLM tool, and translation tool.

Response Format

For non-streaming responses, the search results and answer are in the choices[0].message.content field. For streaming responses, the search results and answer are in the choices[0].delta.content field. The response contains a web_search_data field that provides: