Perplexity API Integration

Sonar API Models

Sonar Pro

Advanced search API with in-depth, multi-step queries. Double the citations and larger context window for complex research tasks.

  • • $5 per 1,000 searches
  • • $3 per ~1M input tokens
  • • $15 per ~1M output tokens
  • • F-score: 0.858 on SimpleQA benchmark

Sonar (Base)

Fast, cost-effective search API using a lightweight model. Most affordable search API product on the market.

  • • $5 per 1,000 searches
  • • $1 per ~1M input tokens
  • • $1 per ~1M output tokens
  • • F-score: 0.773 on SimpleQA benchmark

Sonar Deep Research

Comprehensive research mode for the most thorough analysis and information gathering on complex topics.

  • • Multi-step research process
  • • Comprehensive source analysis
  • • Enterprise-grade research capabilities
  • • Advanced query processing

Search Modes

Choose between High, Medium, and Low search modes to balance performance and pricing for your specific needs.

  • • High: Maximum depth and context
  • • Medium: Balanced approach
  • • Low: Cost-optimized accuracy
  • • Available across all models

⚠️ Important Notice

Information about models, pricing, and features may be outdated or incorrect. Always consult the official provider documentation for the most current and accurate data.

Key Features

  • • Fastest, cheapest grounded search APIs available
  • • Real-time web-wide research and Q&A capabilities
  • • 1200 tokens per second powered by Cerebras infrastructure
  • • Citations returned by default in all requests
  • • Custom source filtering and selection
  • • No LLM training on your data
  • • Outperforms GPT-4o on factual correctness (SimpleQA)
  • • Company valued at $9 billion (Dec 2024)