Parallel AI

Use Parallel AI to search the web, extract content from URLs, or start a deep-research task. Search accepts an objective and query controls, while extraction accepts the target URLs. Set the available result and content limits to keep downstream inputs bounded, and connect a Parallel API key.

Use the block

Search the web, extract information from URLs, and conduct deep research.

Actions

Parallel AI Search

Search the web with Parallel AI and return matching pages with content excerpts.

Input

Parameter	Type	Required	Description
`search_queries`	string	Yes	Comma-separated list of concise keyword search queries (3-6 words each). At least one is required
`objective`	string	No	Natural-language description of the search intent used to rank and excerpt results
`mode`	string	No	Search mode: turbo, fast, basic, or advanced (default: advanced)
`max_results`	number	No	Maximum number of results to return (default: 10, max: 20)
`max_chars_per_result`	number	No	Maximum characters of excerpts per result
`include_domains`	string	No	Comma-separated list of domains to restrict search results to
`exclude_domains`	string	No	Comma-separated list of domains to exclude from search results
`apiKey`	string	Yes	Parallel AI API Key

Output

Parameter	Type	Description
`search_id`	string	Unique identifier for this search request
`results`	array	Search results with excerpts from relevant pages
↳ `url`	string	The URL of the search result
↳ `title`	string	The title of the search result
↳ `publish_date`	string	Publication date of the page (YYYY-MM-DD)
↳ `excerpts`	array	LLM-optimized excerpts from the page

Parallel AI Extract

Extract targeted information from specific URLs using Parallel AI. Processes provided URLs to pull relevant content based on your objective.

Input

Parameter	Type	Required	Description
`urls`	string	Yes	Comma-separated list of URLs to extract information from (up to 20)
`objective`	string	No	What information to extract from the provided URLs (up to 5,000 characters)
`full_content`	boolean	No	Include full page content as markdown in addition to excerpts (default: false)
`apiKey`	string	Yes	Parallel AI API Key

Output

Parameter	Type	Description
`extract_id`	string	Unique identifier for this extraction request
`results`	array	Extracted information from the provided URLs
↳ `url`	string	The source URL
↳ `title`	string	The title of the page
↳ `publish_date`	string	Publication date (YYYY-MM-DD)
↳ `excerpts`	array	Relevant text excerpts in markdown
↳ `full_content`	string	Full page content as markdown (only when requested)
`errors`	array	URLs that could not be extracted, with the reason
↳ `url`	string	The URL that failed
↳ `error_type`	string	Category of the failure
↳ `http_status_code`	number	HTTP status returned by the page, if any
↳ `content`	string	Error detail

Parallel AI Deep Research

Research the web with Parallel AI and synthesize findings from multiple sources with citations. Can take up to 45 minutes to complete.

Input

Parameter	Type	Required	Description
`input`	string	Yes	Research query or question (up to 15,000 characters)
`output_format`	string	No	Output format: text for a markdown report with inline citations, auto for a structured JSON object that needs the pro tier or higher (default: text)
`processor`	string	No	Processing tier: core, core2x, pro, ultra, ultra2x, ultra4x, ultra8x, or a -fast variant (default: pro)
`include_domains`	string	No	Comma-separated list of domains to restrict research to (source policy)
`exclude_domains`	string	No	Comma-separated list of domains to exclude from research (source policy)
`apiKey`	string	Yes	Parallel AI API Key

Output

Parameter	Type	Description
`status`	string	Task status (completed, failed, running)
`run_id`	string	Unique ID for this research task
`message`	string	Status message
`content`	json	Research findings: a markdown report string for the text output format, or a structured object with query-specific keys for the auto format
`basis`	array	Citations and sources with reasoning and confidence levels
↳ `field`	string	Output field dot-notation path
↳ `reasoning`	string	Explanation for the result
↳ `citations`	array	Array of sources
↳ `url`	string	Source URL
↳ `title`	string	Source title
↳ `excerpts`	array	Relevant excerpts from the source
↳ `confidence`	string	Confidence level (low, medium, high)