> For the complete documentation index, see [llms.txt](https://docs.agenticflow.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.agenticflow.ai/reference/nodes/research_deep_research.md).

# Deep Research

**Action ID:** `research_deep_research`

## Description

Research a topic using a deep research agent.

## Connection

| Name              | Description                               | Required | Category |
| ----------------- | ----------------------------------------- | -------- | -------- |
| OpenAI Connection | The OpenAI connection to call OpenAI API. | True     | openai   |

## Input Parameters

| Name  | Type   | Required | Default | Description            |
| ----- | ------ | :------: | ------- | ---------------------- |
| topic | string |     ✓    | -       | The topic to research. |

<details>

<summary>View JSON Schema</summary>

```json
{
  "description": "Deep Research node input.",
  "properties": {
    "topic": {
      "description": "The topic to research.",
      "title": "Topic",
      "type": "string"
    }
  },
  "required": [
    "topic"
  ],
  "title": "DeepResearchInput",
  "type": "object"
}
```

</details>

## Output Parameters

| Name | Type   | Description                                          |
| ---- | ------ | ---------------------------------------------------- |
| data | object | The research data containing comprehensive findings. |

<details>

<summary>View JSON Schema</summary>

```json
{
  "description": "Deep Research node output.",
  "properties": {
    "data": {
      "additionalProperties": true,
      "description": "The research data.",
      "title": "Research Data",
      "type": "object"
    }
  },
  "required": [
    "data"
  ],
  "title": "DeepResearchOutput",
  "type": "object"
}
```

</details>

## How It Works

This node leverages an AI-powered deep research agent that conducts comprehensive research on your specified topic. The agent formulates research questions, searches multiple sources, synthesizes information, and generates a structured report with findings, insights, and relevant data points. It uses OpenAI's capabilities to analyze, summarize, and organize information from various perspectives.

## Usage Examples

### Example 1: Technology Research

**Input:**

```
topic: "The impact of quantum computing on cryptography"
```

**Output:**

```
data: {
  "summary": "Quantum computing poses significant threats to current cryptographic systems...",
  "key_findings": [
    "RSA and ECC encryption vulnerable to Shor's algorithm",
    "Post-quantum cryptography standards being developed by NIST",
    "Timeline: 10-15 years before quantum computers threaten current encryption"
  ],
  "sources": [...],
  "recommendations": "Organizations should begin planning quantum-resistant strategies..."
}
```

### Example 2: Market Research

**Input:**

```
topic: "Electric vehicle market trends in Europe 2024"
```

**Output:**

```
data: {
  "summary": "The European EV market shows strong growth with 25% year-over-year increase...",
  "key_findings": [
    "Norway leads with 87% EV market share",
    "Battery costs decreased 40% since 2020",
    "Government incentives driving adoption"
  ],
  "market_size": "2.3 million units sold in 2024",
  "growth_projections": {...}
}
```

### Example 3: Academic Research

**Input:**

```
topic: "Recent advances in CRISPR gene editing technology"
```

**Output:**

```
data: {
  "summary": "CRISPR technology has advanced significantly with new base editing capabilities...",
  "key_findings": [
    "Prime editing allows precise DNA modifications",
    "Clinical trials show promise for sickle cell disease",
    "Ethical considerations remain paramount"
  ],
  "publications": [...],
  "future_directions": "Expanded applications in agriculture and medicine..."
}
```

## Common Use Cases

* **Market Analysis**: Research industry trends, competitor analysis, and market opportunities
* **Academic Research**: Gather information on scientific topics, literature reviews, and research summaries
* **Business Intelligence**: Investigate companies, technologies, or business strategies
* **Policy Research**: Analyze policy implications, regulatory landscapes, and compliance requirements
* **Technology Assessment**: Evaluate emerging technologies, their capabilities, and potential impacts
* **Competitive Intelligence**: Research competitors, their products, and market positioning
* **Trend Analysis**: Identify and analyze trends in various domains like fashion, technology, or consumer behavior

## Error Handling

| Error Type           | Cause                                   | Solution                                                                  |
| -------------------- | --------------------------------------- | ------------------------------------------------------------------------- |
| Authentication Error | Invalid or missing OpenAI API key       | Verify your OpenAI connection is properly configured with a valid API key |
| Empty Topic          | Topic parameter is empty or null        | Provide a clear, specific research topic                                  |
| Rate Limit Error     | Too many API requests in a short period | Implement delays between requests or upgrade your OpenAI plan             |
| Timeout Error        | Research taking too long to complete    | Simplify the topic or break it into smaller research questions            |
| Insufficient Context | Topic is too vague or broad             | Provide more specific research parameters and constraints                 |
| API Error            | OpenAI service issues                   | Check OpenAI status page and retry after a brief delay                    |

## Notes

* **Topic Specificity**: Provide clear, specific topics for better research results. Vague topics may produce generic findings.
* **Research Depth**: The agent conducts multi-step research, which may take longer than simple API calls. Allow sufficient time for completion.
* **Data Quality**: Results depend on the information available and the AI model's training data. Always verify critical information.
* **Token Usage**: Deep research consumes significant tokens due to multiple API calls. Monitor your OpenAI usage and costs.
* **Structured Output**: The data object contains structured findings that can be easily parsed and used in subsequent workflow nodes.
* **Source Verification**: While the agent synthesizes information, always cross-reference important findings with original sources when possible.
* **Cost Optimization**: Use deep research for complex topics that require comprehensive analysis. For simple questions, consider using standard AI nodes instead.