ReasoningPerplexity

Sonar Deep Research

Sonar Deep Research is Perplexity's reasoning-tier model designed for in-depth research and analysis tasks, featuring a 128K token context window.

Context 128K
Tier Reasoning
Input from
$2.00 / 1M tokens
across 2 providers

API Pricing

ProviderInput / 1MOutput / 1MUpdated
$2.00$8.004/13/2026
$2.00$8.004/14/2026

Prices updated daily. Last check: 4/14/2026

Model Details

General

Creator
Perplexity
Family
Sonar
Tier
Reasoning
Context Window
128K
Modalities
Text

Capabilities

Tool Calling
No
Open Source
No

Strengths & Limitations

  • 128K token context window enables processing of lengthy research documents
  • Reasoning-tier specialization for in-depth analytical tasks
  • Part of Perplexity's research-focused model ecosystem
  • Text-focused design optimized for document analysis
  • Designed for comprehensive exploration rather than quick responses
  • No tool calling capabilities for external integrations
  • Text-only modality with no image or multimodal support
  • Proprietary model with no open-source availability
  • Smaller context window compared to some frontier models
  • Limited to reasoning tasks rather than general-purpose use

Key Features

128K token context window
Text input and output processing
Reasoning-tier analytical capabilities
Deep research methodology
Comprehensive topic exploration
Extended document analysis support

About Sonar Deep Research

Sonar Deep Research is Perplexity's reasoning-tier model within the Sonar family, positioned as a specialized tool for comprehensive research and analytical tasks. As a proprietary model from Perplexity, it represents the company's focus on research-oriented AI capabilities rather than general-purpose conversation. The model operates with a 128K token context window and handles text-only inputs, without support for tool calling or multimodal capabilities. Its design emphasizes deep analytical reasoning over breadth of features, targeting users who need thorough examination of complex topics rather than quick responses or diverse functionality. Sonar Deep Research is primarily used for research workflows, academic analysis, and situations requiring comprehensive exploration of subjects. Unlike general-purpose models that balance speed and capability, this model prioritizes depth and thoroughness in its reasoning processes, making it suitable for users who value detailed analysis over rapid iteration.

Common Use Cases

Sonar Deep Research is designed for users who need comprehensive analysis and in-depth exploration of complex topics. Its reasoning-tier capabilities make it suitable for academic research, detailed document analysis, thorough investigation of multifaceted subjects, and situations where depth of analysis is more important than speed of response. The 128K context window supports lengthy research documents and extensive source materials, making it appropriate for literature reviews, policy analysis, and detailed comparative studies where thorough examination takes precedence over rapid iteration.

Frequently Asked Questions

How much does Sonar Deep Research cost per million tokens?

Sonar Deep Research pricing varies by provider and usage patterns. Check the pricing table above for current rates across all available providers offering this model.

What is Sonar Deep Research best used for?

Sonar Deep Research excels at comprehensive analysis and in-depth research tasks. Its reasoning-tier design makes it suitable for academic research, detailed document analysis, thorough investigation of complex topics, and situations where analytical depth is more important than response speed.

Does Sonar Deep Research support tool calling or multimodal inputs?

No, Sonar Deep Research is text-only and does not support tool calling or multimodal inputs like images. It focuses specifically on text-based reasoning and research analysis within its 128K token context window.