Version: 1.12.x (Next)
Exa
tip
If you installed lfx directly (not as part of langflow), install the Exa bundle separately:
- Run
uv pip install lfx-exa. - Restart Langflow.
- To confirm the bundle loaded, run
lfx extension list.
If you installed Langflow with uv pip install langflow, these bundle components are already included.
For more information, see Install LFX with bundle components.
Bundles contain custom components that support specific third-party integrations with Langflow.
This page describes the components that are available in the Exa bundle.
Exa Search
This component provides an Exa toolkit for search and content retrieval, for use by a Langflow Agent component or an MCP client.
It exposes two tools:
search— search the web with Exa and return results, including highlights by default.get_contents— fetch highlights and/or full text for one or more result IDs returned bysearch.
Get an API key from the Exa dashboard. For details on the underlying API and best practices, see the Exa Search API guide.
The output is exclusively Tools.
Exa Search parameters
| Name | Type | Description |
|---|---|---|
Exa API Key (exa_api_key) | SecretString | Input parameter. Your Exa API key. |
Search type (search_type) | Dropdown | Input parameter. One of auto, fast, instant, or deep. auto is the default and recommended for most use cases. |
Number of results (search_num_results) | Integer | Input parameter. Maximum number of results to return for search. Default: 10. |
Include highlights (include_highlights) | Boolean | Input parameter. Return token-efficient highlights with each result. Default: true. |
Highlights max characters (highlights_max_characters) | Integer | Input parameter. Optional cap on the length of each highlight, in characters. 0 leaves it at Exa's default. |
Include full text (include_text) | Boolean | Input parameter. Return full page text. Default: false. Prefer highlights for token efficiency. |
Category (category) | Dropdown | Input parameter. Optional. Restrict results to one of company, people, research paper, news, personal site, or financial report. |
Max age (hours) (max_age_hours) | Integer | Input parameter. Optional. Max age of cached content (in hours) before refetching. 0 = always refetch. -1 = never refetch (use cache only). |
Include domains (include_domains) | String | Input parameter. Optional. Comma-separated allowlist of domains. |
Exclude domains (exclude_domains) | String | Input parameter. Optional. Comma-separated denylist of domains. |
Start published date (start_published_date) | String | Input parameter. Optional. ISO date (YYYY-MM-DD). Only return results published on or after this date. |
End published date (end_published_date) | String | Input parameter. Optional. ISO date (YYYY-MM-DD). Only return results published on or before this date. |
Was this page helpful?