Back to models
Perplexity logo

Perplexity

Sonar

Perplexity’s lightweight search model for quick grounded answers with real-time web search.

Overall score
85
searchgroundedweb
Context window
Search-context dependent
Speed
Fast
Input pricing
$1 / 1M input tokens
Output pricing
$1 / 1M output tokens

Score breakdown

Capability80
Use-case fit90
Cost efficiency91
Speed94
Reliability86
Agent readiness78
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

Search-backed answersnews monitoringweb-grounded assistants

Modalities

textsearch

Sources & trust

Officially verified core fields
Official linkSummaryDescriptionTool supportPricingPricing page

Editorial fields such as shortlist guidance, strengths, caveats, and scoring remain clearly separated from official provider data.

Last verified: Apr 9, 2026

Strengths

  • Fast grounded answers are the core product strength.
  • Good for teams that care more about answer freshness than raw frontier reasoning.

Things to watch

  • This is a search-native product tier, not a general-purpose long-context flagship.
  • Request-fee economics should be reviewed alongside token prices.

Best for