Text Generation Model Deprecated
Sonar Reasoning
Lightweight reasoning offering powered by reasoning models trained with DeepSeek R1.
Publisher
Perplexity
Type Text
Context Window 127,000 tokens
Training Data November 2023
Input $1.00/MTok
Output $5.00/MTok
Overview
Sonar Reasoning
Chain-of-thought reasoning, plus real-time internet search and citations. Sonar Reasoning is uncensored and hosted in US datacenters.
Ready to build with Sonar Reasoning?
Get Started FreeConfiguration
Parameters & options
Max Temperature 1.9
Max Response Size 32,768 tokens
Return Citations Select
Determines whether or not a request to an online model should return citations.
Return Images Select
Determines whether or not a request to an online model should return images.
Search Context Size Select
Controls how much web information is retrieved. Higher context provides more comprehensive results but costs more per request.
Related models
Explore similar models
Start building with Sonar Reasoning
No API keys required. Create AI-powered workflows with Sonar Reasoning in minutes — free.