Search & Research
⚙Tool
Parallel AI
Web search, URL extraction, and deep research with citations via Parallel AI
Web research powered by Parallel AI. Search the web with intelligent processing, extract targeted information from specific URLs, and run deep research that synthesizes multiple sources with citations.
Overview
| Property | Value |
|---|---|
| Type | parallel_ai |
| Category | Tool — Search |
| Auth | API Key (PARALLEL_API_KEY or the block's API Key field) |
Operations
| Operation | Tool ID | Description |
|---|---|---|
| Search | parallel_search | Search the web against an objective, optionally with specific queries, mode, and domain filters. |
| Extract from URLs | parallel_extract | Pull excerpts and/or full content from a list of URLs against an objective. |
| Deep Research | parallel_deep_research | Run a long-form research task (up to ~45 min) returning structured content plus a basis of citations with confidence. |
Configuration
- Operation — choose Search, Extract, or Deep Research.
- API Key — your Parallel AI API key (sent as the
x-api-keyheader). - Search:
Search Objective(required), optionalSearch Queries(comma-separated), advancedSearch Mode(one-shot/agentic/fast),Max Results,Max Chars Per Result, and include/exclude domain filters. - Extract:
URLs(comma-separated, required), optionalExtract Objective,Include Excerpts(default yes),Include Full Content(default no). - Deep Research:
Research Query(required), advancedResearch Processor(lite/base/core/pro/ultra), and include/exclude domain filters.
Outputs
- Search / Extract:
results(array ofurl,title,publish_date,excerpts, andfull_contentfor extract) plussearch_id/extract_id. - Deep Research:
status,run_id,message,content(structured research output), andbasis(per-field reasoning, citations, and confidence).
YAML Example
- type: parallel_ai
name: Research company
inputs:
operation: deep_research
research_input: "Recent funding, leadership changes, and product launches for Acme Corp"
processor: pro
apiKey: "{{PARALLEL_API_KEY}}"