> 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/perplexity_search.md).

# Perplexity Search

**Action ID:** `perplexity_search`

## Description

Search the web and get AI-powered answers using Perplexity's advanced search models. This node provides real-time information retrieval combined with AI reasoning to deliver accurate, cited answers to your queries.

## Provider

**Perplexity**

## Connection

| Name                  | Description                                      | Required | Category   |
| --------------------- | ------------------------------------------------ | :------: | ---------- |
| Perplexity Connection | The Perplexity connection to use for the search. |     ✓    | perplexity |

## Input Parameters

| Name        | Type   | Required | Default | Description                                                                                                                                |
| ----------- | ------ | :------: | ------- | ------------------------------------------------------------------------------------------------------------------------------------------ |
| prompt      | string |     ✓    | -       | The prompt to send to Perplexity                                                                                                           |
| model       | string |     -    | sonar   | The model to use for chat completion. Available options: sonar, sonar-pro                                                                  |
| temperature | number |     -    | 0.7     | Controls randomness in the response. Higher values make the output more random, lower values make it more deterministic. Range: 0.0 to 1.0 |

<details>

<summary>View JSON Schema</summary>

```json
{
  "description": "Perplexity search node input.",
  "properties": {
    "prompt": {
      "title": "Prompt",
      "type": "string",
      "description": "The prompt to send to Perplexity."
    },
    "model": {
      "title": "Model",
      "type": "string",
      "enum": ["sonar", "sonar-pro"],
      "default": "sonar",
      "description": "The model to use for chat completion."
    },
    "temperature": {
      "title": "Temperature",
      "type": "number",
      "default": 0.7,
      "minimum": 0.0,
      "maximum": 1.0,
      "description": "Controls randomness in the response. Higher values make the output more random, lower values make it more deterministic."
    }
  },
  "required": [
    "prompt"
  ],
  "title": "PerplexitySearchInput",
  "type": "object"
}
```

</details>

## Output Parameters

| Name     | Type | Description                  |
| -------- | ---- | ---------------------------- |
| response | any  | The response from Perplexity |

<details>

<summary>View JSON Schema</summary>

```json
{
  "description": "Perplexity search node output.",
  "properties": {
    "response": {
      "title": "Response",
      "description": "The response from Perplexity."
    }
  },
  "required": [
    "response"
  ],
  "title": "PerplexitySearchOutput",
  "type": "object"
}
```

</details>

## How It Works

This node sends your prompt to Perplexity's AI search service, which searches the web in real-time for relevant information and processes it through advanced language models to generate accurate, cited answers. The response includes both the AI-generated answer and citations to source materials.

## Usage Examples

### Example 1: Current Events Search

**Input:**

```
prompt: "What are the latest developments in artificial intelligence?"
model: "sonar"
temperature: 0.5
```

**Output:**

```
response: {
  "answer": "Recent developments in AI include...",
  "citations": [
    {"title": "Latest AI Breakthroughs", "url": "https://example.com/ai-news"},
    {"title": "AI Research Updates", "url": "https://example.com/research"}
  ]
}
```

### Example 2: Technical Question with Pro Model

**Input:**

```
prompt: "How do I implement OAuth 2.0 authentication in Python?"
model: "sonar-pro"
temperature: 0.3
```

**Output:**

```
response: {
  "answer": "OAuth 2.0 implementation steps:\n1. Install required libraries...",
  "code_examples": [
    {"language": "python", "code": "..."}
  ],
  "citations": [
    {"title": "OAuth 2.0 Documentation", "url": "https://example.com/oauth"}
  ]
}
```

### Example 3: Product Research

**Input:**

```
prompt: "Compare the best laptop models available in 2024 for programming"
model: "sonar"
temperature: 0.6
```

**Output:**

```
response: {
  "answer": "Top programming laptops in 2024:\n1. MacBook Pro...",
  "comparisons": [
    {"model": "MacBook Pro", "specs": {...}, "price": "$1999"},
    {"model": "ThinkPad", "specs": {...}, "price": "$1299"}
  ],
  "citations": [
    {"title": "Best Laptops 2024", "url": "https://example.com/reviews"}
  ]
}
```

## Common Use Cases

* **Research Assistance**: Get comprehensive, cited answers to research questions
* **Current Events Awareness**: Stay updated on breaking news and latest developments
* **Technical Documentation**: Find and summarize technical solutions and best practices
* **Product Research**: Compare products and gather pricing and specification information
* **Market Intelligence**: Research competitor products and industry trends
* **Academic Research**: Find scholarly articles and research papers on specific topics
* **Real-time Data**: Get current information that may not be available in training data

## Error Handling

| Error Type           | Cause                                                       | Solution                                                 |
| -------------------- | ----------------------------------------------------------- | -------------------------------------------------------- |
| Authentication Error | Invalid or missing Perplexity API key                       | Verify your Perplexity connection is properly configured |
| Invalid Model        | Model name doesn't exist or access not granted              | Use only 'sonar' or 'sonar-pro' models                   |
| Network Error        | Cannot reach Perplexity servers or search the web           | Check internet connection and Perplexity service status  |
| Rate Limit Exceeded  | Too many requests in a short period                         | Implement delays between requests or upgrade your plan   |
| Empty Response       | Search returned no results or answer could not be generated | Rephrase the prompt to be more specific                  |
| Timeout              | Request took too long to process                            | Simplify the prompt or try again with reduced complexity |

## Notes

* **Model Selection**: Choose 'sonar' for standard searches and 'sonar-pro' for more complex analysis requiring deep reasoning.
* **Temperature Control**: Lower temperature (0.0-0.3) for factual, deterministic responses. Higher temperature (0.7-1.0) for more creative exploration.
* **Citations**: Perplexity includes source citations for transparency and fact-checking.
* **Real-time Search**: This node searches the current web, so it provides up-to-date information beyond training data cutoffs.
* **Specific Prompts**: More specific and detailed prompts yield better results. Include context and any constraints.
* **Cost Optimization**: The standard 'sonar' model is cost-effective for most searches. Reserve 'sonar-pro' for complex queries.