All AI models · LLM benchmarks · Methodology

Perplexity Sonar Reasoning

Web-grounded reasoning model for multi-step research and cited answers

Model facts

Context window
127K tokens
Maximum output
4K tokens
Cheapest paid input
$1 per 1M tokens
Cheapest paid output
$5 per 1M tokens
Open weights
no
Providers
1

Capabilities and modalities

reasoning, temperature control, text.

Observed price history

  • 2026-08-24: $1 input / $5 output per 1M tokens

API providers