Exa

Reference for the Tracecat Exa integration: registered actions, required secrets, expected inputs, and example workflow usage.

Answer

Action ID: tools.exa.answer

Answer a question with Exa search-grounded citations.

Reference: https://exa.ai/docs/reference/answer

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

query · string · required

Natural-language question or instructions for the request.

outputSchema · object | null

JSON Schema Draft 7 object for structured answer output.

Default: null.

text · boolean | null

Return full page text in citations when true.

Default: null.

Deep research

Action ID: tools.exa.deep_research

Run Exa deep-reasoning search for research-style discovery and synthesis.

Reference: https://exa.ai/docs/changelog/may-2026-api-deprecations

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

query · string · required

Natural language research query or instructions.

additionalQueries · array[string] | null

Additional query variations used alongside the main query.

Default: null.

compliance · string | null

Optional Exa enterprise compliance mode.

Default: null.

contents · object | null

Optional Exa contents options nested under the search request contents field.

Default: null.

numResults · integer | null

Number of results to return. Exa supports 1 to 100.

Default: null.

outputSchema · object | null

JSON schema for synthesized output. Supported root types are text and object.

Default: null.

systemPrompt · string | null

Additional instructions that guide synthesized output or search planning.

Default: null.

Get contents

Action ID: tools.exa.get_contents

Get clean page contents, highlights, summaries, and metadata for URLs or Exa result IDs.

Reference: https://exa.ai/docs/reference/contents-api-guide-for-coding-agents

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

urls · array[string] · required

URLs or Exa result IDs to extract content from.

compliance · string | null

Optional Exa enterprise compliance mode.

Default: null.

extras · object | null

Optional Exa extras options, such as links or imageLinks.

Default: null.

highlights · boolean | object | null

Return extractive key excerpts, or Exa highlight options.

Default: null.

livecrawlTimeout · integer | null

Timeout for livecrawling in milliseconds.

Default: null.

maxAgeHours · integer | null

Maximum age of cached content in hours. Use 0 to always livecrawl or -1 for cache-only.

Default: null.

subpageTarget · string | array[string] | null

Keywords to prioritize when selecting subpages.

Default: null.

subpages · integer | null

Number of subpages to crawl for each URL.

Default: null.

summary · boolean | object | null

Return LLM-generated summaries, or Exa summary options.

Default: null.

text · boolean | object | null

Return full page text as markdown, or Exa text options.

Default: true.

Search companies

Action ID: tools.exa.search_companies

Search Exa’s company index for company profiles and typed company metadata as a second opinion for company-enrichment workflows.

Reference: https://exa.ai/docs/reference/verticals/company-for-coding-agents

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

query · string · required

Natural language company search query. Encode industry, funding stage, headcount, geography, technology, similarity, and timing constraints in the query string.

contents · object | null

Optional Exa contents options nested under the search request contents field, such as highlights, text, or summary.

Default: null.

numResults · integer | null

Number of company results to return. Exa supports 1 to 100.

Default: null.

outputSchema · object | null

Optional JSON schema for structured output when using deep search modes.

Default: null.

type · string | null

Search method. Exa recommends auto; deep and deep-reasoning also work.

Default: "auto".

Search news

Action ID: tools.exa.search_news

Search Exa’s news index for current events, journalism, campaign chatter, and executive-protection monitoring.

Reference: https://exa.ai/docs/reference/verticals/news-for-coding-agents

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

query · string · required

Natural language news query for a topic, company, person, event, domain, product, vulnerability, or campaign artifact.

compliance · string | null

Optional Exa enterprise compliance mode.

Default: null.

contents · object | null

Optional Exa contents options nested under the search request contents field, such as highlights, text, summary, maxAgeHours, subpages, or extras.

Default: null.

endPublishedDate · string | null

ISO 8601 upper bound for published date.

Default: null.

excludeDomains · array[string] | null

Exclude news results from these domains.

Default: null.

includeDomains · array[string] | null

Only return news results from these domains.

Default: null.

numResults · integer | null

Number of news results to return. Exa supports 1 to 100.

Default: null.

outputSchema · object | null

Optional JSON schema for structured output when using Exa synthesis.

Default: null.

startPublishedDate · string | null

ISO 8601 lower bound for published date.

Default: null.

type · string | null

Search method. Exa recommends auto for news; use fast or instant when latency matters.

Default: "auto".

Search people

Action ID: tools.exa.search_people

Search Exa’s people index for professional profiles as a second opinion for people-enrichment workflows.

Reference: https://exa.ai/docs/reference/verticals/people-for-coding-agents

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

query · string · required

Natural language people search query. Encode role, skill, company, location, seniority, and other constraints in the query string.

contents · object | null

Optional Exa contents options nested under the search request contents field, such as highlights, text, or summary.

Default: null.

numResults · integer | null

Number of people results to return. Exa supports 1 to 100.

Default: null.

outputSchema · object | null

Optional JSON schema for structured output when using deep search modes.

Default: null.

type · string | null

Search method. Exa recommends auto; deep and deep-reasoning also work.

Default: "auto".

Search the web

Action ID: tools.exa.search

Search the web with Exa, with optional advanced filters and content extraction.

Reference: https://exa.ai/docs/reference/search-api-guide-for-coding-agents

Secrets

Required secrets:

• exa: required values EXA_API_KEY.

Input fields

query · string · required

Natural language search query. Supports long, semantically rich descriptions.

additionalQueries · array[string] | null

Additional query variations for deep-search variants.

Default: null.

category · string | null

Optional content category. Known categories include company, people, research paper, news, personal site, and financial report; Exa also accepts other category hints.

Default: null.

compliance · string | null

Optional Exa enterprise compliance mode.

Default: null.

contents · object | null

Optional Exa contents options nested under the search request contents field, such as highlights, text, summary, maxAgeHours, subpages, or extras.

Default: null.

endPublishedDate · string | null

ISO 8601 upper bound for published date. Not supported for company or people category searches.

Default: null.

excludeDomains · array[string] | null

Exclude results from these domains. Not supported for company or people category searches.

Default: null.

includeDomains · array[string] | null

Only return results from these domains. Exa supports domain path filters and subdomain wildcards.

Default: null.

moderation · boolean | null

Filter unsafe content from search results.

Default: null.

numResults · integer | null

Number of results to return. Exa supports 1 to 100.

Default: null.

outputSchema · object | null

JSON schema for synthesized output. Supported root types are text and object.

Default: null.

startPublishedDate · string | null

ISO 8601 lower bound for published date. Not supported for company or people category searches.

Default: null.

systemPrompt · string | null

Additional instructions that guide synthesized output or deep-search planning.

Default: null.

type · string | null

Search method. Exa supports auto, instant, fast, deep-lite, deep, and deep-reasoning.

Default: "auto".

userLocation · string | null

Two-letter ISO country code used to bias search results.

Default: null.