Back to models
Perplexity logo

Perplexity

Sonar Reasoning Pro

Research-first model experience optimized for answer grounding, current web synthesis, and citation-friendly output.

Overall score
83
researchcitationsweb
Context window
128K
Speed
Balanced
Input pricing
$2 / 1M tok
Output pricing
$8 / 1M tok

Score breakdown

Capability85
Use-case fit89
Cost efficiency82
Speed80
Reliability84
Agent readiness74
Ecosystem76

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

citation-heavy briefsmarket scansanswer-grounding workflows

Modalities

textweb

Strengths

  • Great for research UX with source visibility
  • Useful for analyst workflows that need current context

Things to watch

  • Not a primary choice for tool-heavy coding agents
  • Less control than fully orchestrated platform APIs

Best for