Portkey AI Gateway with Perplexity

## Overview

Portkey provides an AI Gateway that routes requests to Perplexity's Sonar models with added observability, prompt management, and reliability features. Supports both sonar and sonar-pro models with 127K token context.

## Key Features

- **Citation Retrieval**: Get source citations by disabling strict OpenAI compliance
- **Domain Filtering**: Whitelist or blacklist specific domains for citations
- **Recency Controls**: Filter by month, week, day, or hour for fresh results
- **Search Context Size**: Control depth with low, medium, or high settings
- **Image Results**: Retrieve relevant images alongside text responses (beta)
- **Related Questions**: Generate follow-up question suggestions (beta)

## Supported Models

- sonar: 127K context, real-time search
- sonar-pro: 127K context, advanced reasoning

## Integration Methods

- Python SDK
- TypeScript/JavaScript SDK
- OpenAI-compatible client
- cURL requests
- Responses API format

## Setup

1. Access Portkey dashboard → Model Catalog → Add Provider
2. Select Perplexity AI, enter API key
3. Name the provider (e.g., "perplexity-ai")
4. Use provider name in API calls

Includes observability dashboards, cost tracking, and fallback routing.