Perplexity API Complete Setup Guide: API Key, Python SDK, Citation Parsing & Search Models

Perplexity API Complete Setup Guide

Perplexity AI offers a powerful API that combines large language models with real-time web search capabilities. This guide walks you through everything from obtaining your API key to parsing source citations and selecting the right search-augmented model for your use case.

Step 1: Create a Perplexity Account and Generate Your API Key

• Visit perplexity.ai and sign up for an account (or log in if you already have one).- Navigate to Settings → API or go directly to the API settings page.- Click Generate API Key. Copy the key immediately — it will only be displayed once.- Add billing information. Perplexity API uses a pay-per-request pricing model. You must have a valid payment method on file before making API calls.- Store your key securely. Never commit it to version control.

Store Your Key as an Environment Variable

# Linux / macOS export PERPLEXITY_API_KEY=“YOUR_API_KEY”

echo “PERPLEXITY_API_KEY=YOUR_API_KEY” >> .env

Step 2: Install the Required Python Packages

Perplexity's API is compatible with the OpenAI SDK, so you can use the official openai Python library as your client. # Install the OpenAI Python SDK pip install openai

Optional: for environment variable management

pip install httpx

Step 3: Configure the Python Client

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ.get("PERPLEXITY_API_KEY", "YOUR_API_KEY"),
    base_url="https://api.perplexity.ai"
)

That's it — three lines and you have a fully configured client. The key difference from standard OpenAI usage is the base_url parameter pointing to Perplexity's endpoint.

Step 4: Make Your First API Call

response = client.chat.completions.create( model=“sonar-pro”, messages=[ {“role”: “system”, “content”: “You are a helpful research assistant. Be concise and cite sources.”}, {“role”: “user”, “content”: “What are the latest developments in quantum computing in 2026?”} ] )

print(response.choices[0].message.content)

Step 5: Parse Source Citations

One of Perplexity's most valuable features is returning source citations alongside generated answers. Citations are returned in the response object. response = client.chat.completions.create( model="sonar-pro", messages=[ {"role": "user", "content": "What is retrieval-augmented generation?"} ] )

Access the answer

if hasattr(response, ‘citations’): citations = response.citations print(“\nSources:”) for i, citation in enumerate(citations, 1): print(f” [{i}] {citation}”)

Building a Citation Formatter

import re

def format_response_with_citations(response):
    """Format Perplexity response with numbered source citations."""
    content = response.choices[0].message.content
    citations = getattr(response, 'citations', [])

    if not citations:
        return content

    formatted = content + "\n\n--- Sources ---\n"
    for i, url in enumerate(citations, 1):
        formatted += f"[{i}] {url}\n"

    return formatted

# Usage
result = format_response_with_citations(response)
print(result)

Step 6: Choose the Right Search-Augmented Model

Perplexity offers several models optimized for different use cases:

# Fine-tune search behavior with additional parameters
response = client.chat.completions.create(
    model="sonar-pro",
    messages=[
        {"role": "user", "content": "Latest Python 3.14 features"}
    ],
    temperature=0.2,           # Lower = more factual
    max_tokens=1024,           # Control response length
    top_p=0.9,                 # Nucleus sampling
    search_recency_filter="week",  # Filter: day, week, month, year
    return_related_questions=True   # Get follow-up suggestions
)

# Robust error handling template
import time
from openai import OpenAI, APIError, RateLimitError
client = OpenAI(
api_key=os.environ[“PERPLEXITY_API_KEY”],
base_url=“https://api.perplexity.ai”
)
def query_with_retry(prompt, model=“sonar-pro”, max_retries=3):
for attempt in range(max_retries):
try:
response = client.chat.completions.create(
model=model,
messages=[{“role”: “user”, “content”: prompt}]
)
return response
except RateLimitError:
wait = 2 ** attempt
print(f”Rate limited. Retrying in {wait}s…”)
time.sleep(wait)
except APIError as e:
print(f”API Error: {e}”)
raise
raise Exception(“Max retries exceeded”)

Frequently Asked Questions

Is the Perplexity API compatible with the OpenAI Python SDK?

Yes. Perplexity's API follows the OpenAI chat completions format. You simply install the openai package and point the base_url to https://api.perplexity.ai. All standard parameters like temperature, max_tokens, and stream work as expected, with additional Perplexity-specific options like search_recency_filter.

How do I access and parse citations from Perplexity API responses?

Citations are returned as part of the response object. Access them via response.citations, which provides a list of source URLs referenced in the generated answer. You can pair these with bracketed numbers in the response text to build fully attributed outputs for research or content workflows.

Which Perplexity model should I use for my project?

Use sonar for fast, simple lookups where speed and cost matter most. Choose sonar-pro for complex research requiring multiple search passes and deeper analysis. Pick sonar-reasoning or sonar-reasoning-pro for tasks that involve logic, math, or scientific analysis combined with real-time web data. For exhaustive multi-step reports, sonar-deep-research is the most thorough option.