Sonar
Sonar is a lightweight model developed by Perplexity designed for fast and simple question-and-answer tasks. It supports both text and image inputs and provides citations to help users verify information.
Sonar specs and pricing
- Provider: Perplexity
- Released: January 27, 2025
- Context window: 127K tokens
- Max output: 114K tokens
- Input: text, images
- Output: text
- Price: $1.00 input / $1.00 output per 1M tokens
What Sonar is good at
- Integrating question-and-answer features into applications
- Processing text and image data for quick insights
- Tasks requiring verifiable information through citations
To-dos to try with Sonar
- Analyze this image of a technical diagram and explain how the components connect
- Read this document and provide a summary with citations for each key claim
- Answer a series of technical questions based on the provided source materials
FAQ
- What is the primary purpose of the Sonar model? Sonar is designed to provide fast, lightweight, and simple question-and-answer capabilities for developers and companies.
- Can Sonar process visual information? Yes, the model accepts both text and image inputs.
- How much does Sonar cost? Sonar costs $1.00 per million input tokens and $1.00 per million output tokens through OpenRouter. In todo.is a typical agent step costs about 2.94 credits ($1 = 100 credits).
- What is the context window of Sonar? Sonar can read up to 127K tokens at once and write up to 114K tokens in one reply.
- Can I use Sonar without coding? Yes. In todo.is you pick Sonar as your agent's model and write to-dos in plain words, on the web, the Android app, WhatsApp or Telegram. No API key needed: credits are included, or you can bring your own key.
JavaScript is required to use the todo.is app.