Back to models
Perplexity logo

Perplexity

Sonar Reasoning Pro

Perplexity’s reasoning-enhanced search model for harder grounded questions.

Overall score
88
reasoningsearchanalysis
Context window
Search-context dependent
Speed
Balanced
Input pricing
$2 / 1M input tokens
Output pricing
$8 / 1M output tokens

Score breakdown

Capability88
Use-case fit92
Cost efficiency83
Speed82
Reliability88
Agent readiness80
Ecosystem84

Scores combine benchmark signals, product experience, and editorial weighting. Use them as a practical guide, not an absolute truth claim.

Best for

Research

Works with

deeper search workflowsanalysis assistantsresearch copilots

Modalities

textsearch

Strengths

  • Better fit when grounded research also needs explicit reasoning.
  • Price sits below Sonar Pro on token costs.

Things to watch

  • Still inherits search-context pricing complexity.
  • Not positioned as a universal coding or automation default.

Best for