Deep Research API | Unified Agent APIs

Overview

Multi-step deep research: search, scrape, enrich entities, and synthesize into a cited report

Credits

Standard: 10, Deep: 20, Comprehensive: 35

Providers

SerperDevPerplexityFirecrawlApolloClaude

SDK Method

client.research(...)

Modes

The mode parameter is a preset that configures optimal provider routing for common use cases. You can also set individual parameters directly for fine-grained control.

Standard

10 credits

Quick research, up to 5 searches and 8 scrapes.

Fast research for factual questions. 30-60s.

Deep

20 credits

Thorough research with refinement.

Multi-angle research with gap analysis. 1-3 min.

Comprehensive

35 credits

Exhaustive research with iterative refinement.

Maximum depth, up to 30 searches and 40 scrapes. 3-8 min.

Parameters

queryRequired

string

The research question or objective.

depth

select (default: standard)

Research depth.

Standard

Deep

Comprehensive

enrich_entities

boolean (default: true)

Enrich detected people and companies with profile and firmographic data.

include_analysis

boolean (default: false)

Generate computational analysis with charts.

synthesis

boolean (default: true)

Generate a synthesized report. Set false for raw mode (sources + entities only).

source_types

string (default: web)

Comma-separated source types: web, news, academic.

max_sources

number (default: 15)

Maximum number of sources to collect.

freshness

select (default: any)

Time filter for search results.

Any time

Past 24 hours

Past week

Past month

Example Response

{
  "success": true,
  "data": {
    "report": {
      "title": "AI Chip Startups Competing with Nvidia (2026)",
      "summary": "Several startups are challenging Nvidia's dominance in AI accelerators...",
      "sections": [
        {
          "heading": "Market Overview",
          "content": "The AI chip market is projected to reach $120B by 2030 [Source 1]..."
        }
      ],
      "key_findings": [
        "Cerebras and Groq lead in inference speed",
        "Custom silicon adoption growing 40% YoY"
      ],
      "confidence": 0.85
    },
    "sources": [
      {
        "id": 1,
        "url": "https://techcrunch.com/2026/ai-chips",
        "title": "AI Chip Race Heats Up",
        "domain": "techcrunch.com",
        "snippet": "Startups are racing to challenge Nvidia...",
        "source_type": "web"
      }
    ],
    "entities": {
      "people": [
        {
          "name": "Jensen Huang",
          "role": "CEO",
          "company": "Nvidia"
        }
      ],
      "companies": [
        {
          "name": "Cerebras",
          "domain": "cerebras.net",
          "industry": "AI Hardware",
          "employee_count": 500
        }
      ]
    },
    "metadata": {
      "depth_used": "standard",
      "sources_searched": 4,
      "sources_scraped": 5,
      "sources_cited": 5,
      "entities_enriched": 2,
      "refinement_rounds": 0,
      "budget_exhausted": false,
      "total_time_ms": 35000,
      "steps_completed": [
        "query_expansion",
        "web_search",
        "content_extraction",
        "entity_detection",
        "entity_enrichment",
        "synthesis"
      ]
    }
  },
  "metadata": {
    "provider_used": "research_orchestrator",
    "providers_tried": [
      "research_orchestrator"
    ],
    "mode_used": "standard",
    "response_time_ms": 35000,
    "request_id": "req_abc123"
  },
  "credits_used": 10
}

Get Started

Use this API through the O-mega platform. Create an API key in your dashboard, then call the endpoint with your key in the Authorization header.

Try Deep Research

Test Deep Research in the interactive playground. No setup required.

Open Playground