AI-Powered10 creditsMulti-Stage

deep_research

AI-powered multi-stage research that discovers, analyzes, and synthesizes information across multiple sources. Features query expansion, source verification, conflict detection, and comprehensive summaries.

Use Cases

Academic Research

Gather and synthesize information from scholarly articles, papers, and academic sources

Market Research

Analyze market trends, competitor strategies, and industry developments across multiple sources

Due Diligence

Investigate companies, products, or technologies with comprehensive source verification

Fact Checking

Verify claims by cross-referencing multiple sources and detecting conflicts

Content Creation

Research topics comprehensively before writing articles, reports, or documentation

Competitive Intelligence

Track product launches, feature updates, and strategic moves across the industry

Endpoint

POST/api/v1/tools/deep_research

Auth Required

2 req/s on Free plan

10 credits

Parameters

Name	Type	Required	Default	Description
topic	string	Required	-	Research topic or question to investigate Example: Impact of AI on software development productivity
maxDepth	number	Optional	3	Research depth level (1-5): number of recursive source expansions Example: 4
queryExpansion	boolean	Optional	true	Automatically generate related queries to broaden research Example: true
enableConflictDetection	boolean	Optional	true	Identify conflicting information across sources Example: true
maxSources	number	Optional	20	Maximum number of sources to analyze (5-50) Example: 30
sourceTypes	string[]	Optional	["all"]	Filter source types: academic, news, blogs, forums, or all Example: ["academic", "news"]
dateRange	object	Optional	-	Filter sources by publication date Example: {"start": "2023-01-01", "end": "2025-12-31"}

Request Examples

terminalBash

Response Example

200 OK12,450ms

{
  "success": true,
  "data": {
    "topic": "Impact of AI on software development productivity",
    "summary": "Comprehensive research indicates AI tools like GitHub Copilot and ChatGPT have increased developer productivity by 30-55% according to multiple studies...",
    "keyFindings": [
      {
        "finding": "GitHub Copilot users complete tasks 55% faster",
        "sources": [
          "https://github.blog/research",
          "https://arxiv.org/paper1"
        ],
        "confidence": "high"
      },
      {
        "finding": "AI reduces debugging time by 40% on average",
        "sources": [
          "https://research.google/paper",
          "https://stackoverflow.blog/ai-survey"
        ],
        "confidence": "medium"
      }
    ],
    "conflicts": [
      {
        "topic": "Impact on code quality",
        "viewpoint1": {
          "claim": "AI-generated code has 20% more bugs",
          "sources": [
            "https://stanford.edu/paper"
          ]
        },
        "viewpoint2": {
          "claim": "No significant difference in bug rates",
          "sources": [
            "https://mit.edu/study"
          ]
        }
      }
    ],
    "sources": [
      {
        "url": "https://github.blog/research",
        "title": "Research: quantifying GitHub Copilot's impact",
        "type": "research",
        "relevance": 0.95,
        "credibility": 0.92,
        "datePublished": "2024-06-15"
      }
    ],
    "sourcesAnalyzed": 28,
    "queriesExpanded": [
      "AI coding assistants productivity",
      "GitHub Copilot studies",
      "ChatGPT developer efficiency"
    ],
    "confidenceScore": 87,
    "depthReached": 4
  },
  "credits_used": 10,
  "credits_remaining": 990,
  "processing_time": 12450
}

Field Descriptions

data.summaryExecutive summary synthesizing all findings

data.keyFindingsMost important discoveries with source citations

data.conflictsContradictory information found across sources

data.confidenceScoreOverall confidence in research quality (0-100)

data.sourcesAnalyzedTotal number of sources reviewed

credits_usedFixed 10 credits per deep research request

processing_timeResearch takes 10-30 seconds depending on depth

Error Handling

Topic Too Vague (400 Bad Request)

The research topic is too broad. Be more specific for better results.

Insufficient Sources (404 Not Found)

Couldn't find enough sources for the topic. Try broader search terms or different date ranges.

Research Timeout (500 Internal Server Error)

Research took too long. Reduce maxDepth or maxSources parameters.

Insufficient Credits (402 Payment Required)

Deep research requires 10 credits. Add more credits to continue.

Pro Tip: Start with maxDepth=3 for most research. Higher depths (4-5) provide marginal improvements but take significantly longer. Enable conflict detection for controversial or rapidly evolving topics.