Implementation:Langchain ai Langchain ExaSearchTools

Knowledge Sources	Langchain_ai_Langchain
Domains	Tools, Web Search, Exa
Last Updated	2026-02-11 00:00 GMT

Overview

This module provides ExaSearchResults and ExaFindSimilarResults, two LangChain tool implementations that wrap the Exa Search API for query-based search and URL-based similarity search respectively.

Description

The module defines two BaseTool subclasses in the langchain_exa package:

• ExaSearchResults -- A tool that performs Exa searches given a text query and returns search results with content, supporting neural, keyword, and auto search types.
• ExaFindSimilarResults -- A tool that finds pages similar to a given URL using Exa's similarity search, with support for domain filtering, date filtering, and content options.

Both tools validate the Exa API key from the environment and instantiate an Exa client. They support configurable parameters for search refinement including domain filtering, date ranges, content highlighting, summaries, and live crawling.

Usage

Import these tools when building LangChain agents that need web search or similar-page discovery capabilities powered by the Exa API.

Code Reference

Source Location

• Repository: Langchain_ai_Langchain
• File: libs/partners/exa/langchain_exa/tools.py
• Lines: 1-243

Signature

class ExaSearchResults(BaseTool):
    name: str = "exa_search_results_json"
    description: str = "A wrapper around Exa Search. ..."
    client: Exa
    exa_api_key: SecretStr

    def _run(
        self,
        query: str,
        num_results: int = 10,
        text_contents_options: TextContentsOptions | dict[str, Any] | bool | None = None,
        highlights: HighlightsContentsOptions | bool | None = None,
        include_domains: list[str] | None = None,
        exclude_domains: list[str] | None = None,
        start_crawl_date: str | None = None,
        end_crawl_date: str | None = None,
        start_published_date: str | None = None,
        end_published_date: str | None = None,
        use_autoprompt: bool | None = None,
        livecrawl: Literal["always", "fallback", "never"] | None = None,
        summary: bool | dict[str, str] | None = None,
        type: Literal["neural", "keyword", "auto"] | None = None,
        run_manager: CallbackManagerForToolRun | None = None,
    ) -> list[dict] | str: ...


class ExaFindSimilarResults(BaseTool):
    name: str = "exa_find_similar_results_json"
    description: str = "A wrapper around Exa Find Similar. ..."
    client: Exa
    exa_api_key: SecretStr
    exa_base_url: str | None = None

    def _run(
        self,
        url: str,
        num_results: int = 10,
        text_contents_options: TextContentsOptions | dict[str, Any] | bool | None = None,
        highlights: HighlightsContentsOptions | bool | None = None,
        include_domains: list[str] | None = None,
        exclude_domains: list[str] | None = None,
        start_crawl_date: str | None = None,
        end_crawl_date: str | None = None,
        start_published_date: str | None = None,
        end_published_date: str | None = None,
        exclude_source_domain: bool | None = None,
        category: str | None = None,
        livecrawl: Literal["always", "fallback", "never"] | None = None,
        summary: bool | dict[str, str] | None = None,
        run_manager: CallbackManagerForToolRun | None = None,
    ) -> list[dict] | str: ...

Import

from langchain_exa import ExaSearchResults, ExaFindSimilarResults

I/O Contract

Inputs (ExaSearchResults._run)

Name	Type	Required	Description
query	str	Yes	The search query.
num_results	int	No	Number of results to return (1 to 100). Default: 10.
text_contents_options	dict | bool | None	No	How to set page content of results.
highlights	bool | None	No	Whether to include highlights.
include_domains	None	No	Domains to include.
exclude_domains	None	No	Domains to exclude.
start_crawl_date	None	No	Start date for crawl (YYYY-MM-DD).
end_crawl_date	None	No	End date for crawl (YYYY-MM-DD).
start_published_date	None	No	Start publication date (YYYY-MM-DD).
end_published_date	None	No	End publication date (YYYY-MM-DD).
use_autoprompt	None	No	Whether to use autoprompt.
livecrawl	None	No	Live crawl option.
summary	dict[str, str] | None	No	Whether to include summary.
type	None	No	Search type.

Inputs (ExaFindSimilarResults._run)

Name	Type	Required	Description
url	str	Yes	The URL to find similar pages for.
num_results	int	No	Number of results (1 to 100). Default: 10.
exclude_source_domain	None	No	If True, exclude pages from the same domain as the source URL.
category	None	No	Filter for similar pages by category.

Outputs

Name	Type	Description
return	str	Search results as a list of dictionaries, or an error string if an exception occurs.

Usage Examples

Basic Usage

from langchain_exa import ExaSearchResults, ExaFindSimilarResults

# Search tool
search_tool = ExaSearchResults()
results = search_tool.invoke({"query": "latest AI research", "num_results": 5})

# Find similar tool
similar_tool = ExaFindSimilarResults()
results = similar_tool.invoke({
    "url": "https://arxiv.org/abs/2301.00001",
    "num_results": 5,
})

With ToolCall

from langchain_exa import ExaSearchResults

tool = ExaSearchResults()
result = tool.invoke({
    "args": {"query": "what is the weather in SF", "num_results": 1},
    "id": "1",
    "name": tool.name,
    "type": "tool_call",
})

Related Pages

• Requires langchain-exa and exa-py packages