# Perplexity API

## Perplexity API Setup

**Available on: Power plans and higher**

Connect Perplexity to enable automatic research for your articles. Perplexity searches the web in real-time and incorporates accurate, cited information into your content.

***

### Why Perplexity?

Perplexity's Sonar model is a research-focused AI that:

* **Searches the web in real-time** during generation
* **Provides accurate, current information** with citations
* **Eliminates manual research** for each article
* **Improves E-E-A-T signals** with factual content

This is one of Cuppa's most requested features — it puts research on autopilot.

***

### How It Works

When enabled, Cuppa:

1. Takes each section of your article outline (H2-H4 headings)
2. Sends research queries to Perplexity
3. Incorporates relevant facts, statistics, and sources
4. Generates content that's both AI-written and research-backed

The result: articles that sound authoritative because they *are* authoritative.

***

### Step 1: Create a Perplexity Account

1. Go to [perplexity.ai](https://perplexity.ai)
2. Click **Sign Up**
3. Create your account

***

### Step 2: Get API Access

1. Go to [perplexity.ai/settings/api](https://perplexity.ai/settings/api)
2. Click **Generate API Key**
3. Copy your key

Your key looks like: `pplx-xxxxxxxxxxxxxxxxxxxx`

> **Note**: Perplexity API is separate from Perplexity Pro subscription. You'll pay per API call.

***

### Step 3: Add Key to Cuppa

1. Go to **Team Settings** → **API Keys**
2. Find the **Perplexity** section
3. Paste your API key
4. Click **Save**

***

### Step 4: Enable Auto-Research

Turn on Perplexity research in your generation settings:

#### In Single Generation

1. Go to **Create Content** → **Single Generation**
2. Look for **"Auto-fetch Research from Perplexity"**
3. Toggle it **ON**

#### In Presets

1. Go to **Saved Presets** → **My Presets**
2. Create or edit a preset
3. Enable **"Auto-fetch Research from Perplexity"**
4. Save the preset

#### In Bulk Generation

1. Go to **Create Content** → **Bulk Generation**
2. In project settings, enable Perplexity research
3. All articles will include research

***

### Plan Requirements

| Plan     | Perplexity Access |
| -------- | ----------------- |
| Hobby    | ❌ Not available   |
| Power    | ✅ Available       |
| Business | ✅ Available       |
| Agency   | ✅ Available       |
| Agency+  | ✅ Available       |

***

### Cost

Perplexity charges per API request. With Cuppa:

* **\~$0.005-0.02** per section researched
* **\~$0.05-0.20** per full article (depending on outline complexity)

This is separate from your main AI model costs (OpenAI/Claude).

***

### Model Used

Cuppa uses **Perplexity Sonar** — their latest, most capable research model. This is configured automatically; you don't need to select a model.

***

### When to Use Perplexity

#### Great for:

* **Factual content** — statistics, data, recent events
* **Technical articles** — accurate specifications, comparisons
* **News/trends** — current information that changes
* **Expert topics** — medical, legal, financial content
* **Product roundups** — up-to-date pricing and features

#### Maybe skip for:

* **Opinion pieces** — where research isn't needed
* **Personal stories** — based on your experience
* **Creative content** — fiction, poetry
* **Very niche topics** — where web data is limited

***

### Troubleshooting

#### "Invalid API key"

* Check you copied the full key (starts with `pplx-`)
* Verify the key exists in your Perplexity settings
* Generate a new key if needed

#### "Perplexity option not showing"

* Verify you're on Power plan or higher
* Check that your API key is saved correctly
* Refresh the page

#### "Research taking too long"

* Perplexity adds 5-15 seconds per article section
* This is normal — research takes time
* For speed, disable research for drafts

#### "Research seems inaccurate"

* Perplexity searches the web, which may have outdated info
* Review and verify critical facts
* Add corrections in the Cuppa editor

***

### FAQ

**Does Perplexity replace OpenAI?**

No. Perplexity provides research data. OpenAI/Claude generates the actual content. They work together.

**Can I see what Perplexity found?**

Currently, research is incorporated into the content automatically. We're working on showing research sources.

**Is Perplexity Pro the same as Perplexity API?**

No. Perplexity Pro is their subscription product. The API is separate with pay-per-use pricing.

**What happens if Perplexity is down?**

Article generation continues without research. You can regenerate later with research enabled.

**Can I use Perplexity for bulk generation?**

Yes! It works with both single and bulk generation. Just enable it in your settings.

**Does research slow down generation?**

Yes, slightly. Each section adds 5-15 seconds for research. For a 10-section article, expect 1-3 minutes of additional time.

***

### Related Docs

* OpenAI API Setup — Required for all users
* Anthropic API Setup — Add Claude models
* Presets Guide — Save research settings in presets