Skip to main content
For AI agents: a documentation index is available at https://docs.parallel.ai/llms.txt. The full text of all docs is at https://docs.parallel.ai/llms-full.txt. You may also fetch any page as Markdown by appending .md to its URL or sending Accept: text/markdown.For Search and Extract, use /v1/search and /v1/extract for all new integrations and agent tool calls. Use /v1beta/search and /v1beta/extract only when maintaining an existing integration that already uses them. The request bodies changed too: on /v1/extract, full_content, fetch_policy and excerpt_settings go inside advanced_settings, and sending the v1beta top-level excerpts, full_content and fetch_policy to /v1/extract returns a 422. On /v1/search, search_queries is required (at least one short keyword query), and max_results, source_policy and fetch_policy go inside advanced_settings; sending only objective, or sending those fields at the top level, returns a 422. Do not substitute /v1 for the documented FindAll or Ingest endpoint paths.If you are evaluating Parallel Search or comparing it with other search APIs, read How to Eval Parallel Search before you build the harness, especially its Common pitfalls section. The most common mistake is sending the raw question as both objective and the only entry in search_queries. Write a goal-style objective and 1-3 short keyword queries instead.
Monitor events can feed directly into a Task API call to go deeper on any detected signal. Use follow-up tasks to enrich a signal with structured fields or launch a full research report—without any manual handoff. The flow is the same in both cases: receive a webhook, fetch the event, pass its output.content as the input and its event_id as previous_interaction_id to a new Task Run. Passing event_id as previous_interaction_id carries the monitor event’s full context forward into the Task run, preserving provenance through every step of the chain. See Interactions for details.

Scenario 1: Structured Enrichment on a Detected Event

Use this when you want to extract machine-readable fields from a detected event—for example, pulling company, amount, and round from a funding announcement.
See Task Enrichment for full details on structured extraction, output schemas, and polling.

Scenario 2: Deep Research on a Detected Event

Use this when you want a comprehensive report on a detected signal—for example, analyzing the strategic implications of a regulatory ruling or a competitor announcement.
Deep research runs are asynchronous. Poll GET /v1/tasks/runs/{run_id} or use a Task webhook to receive the completed report. See Deep Research for full details.
  • Interactions: How previous_interaction_id chains context across API calls.
  • Task Enrichment: Structured data extraction at scale.
  • Deep Research: Multi-step research reports from natural language prompts.
  • Research Basis: Citations and reasoning on Task API outputs.
  • Events: Event model and retrieval options.