Perplexity AI models

Maker of the Sonar search-grounded models, served via the Perplexity API.

Lifecycle watch · 1

Sonar Reasoning — Retired, shuts down 15 Dec 2025 → migrate to sonar-reasoning-pro

The lineup

All Perplexity models

Model ↕	Status ↕	Context ↕	Input $/M ↕	Output $/M ↕	Blended $/M ↕	Cutoff ↕
Sonar Pro Perplexity · text, image	GA	200K	$3	$15	$6	—
Sonar Perplexity · text, image	GA	128K	$1	$1	$1	—
Sonar Reasoning Pro Perplexity · text	GA	128K	$2	$8	$3.50	—
Sonar Deep Research Perplexity · text	GA	128K	$2	$8	$3.50	—
Sonar Reasoning Perplexity · text	Retired	128K	—	—	—	—

Blended = 0.75 × input + 0.25 × output $/M tokens (a fair single-number cost proxy). Click any header to sort.

FAQ

What is the cheapest Perplexity model?

Sonar is the cheapest generally-available Perplexity model we track, at $1 per 1M input tokens and $1 per 1M output tokens ($1/1M blended).

What is Perplexity's flagship model?

Sonar Pro is Perplexity's most prominent model in our catalog, with a 200K-token context window and pricing of $3/$15 per 1M input/output tokens.

How many Perplexity models are there?

We track 5 Perplexity models, of which 4 are generally available and 1 is deprecated or scheduled for retirement.

Which Perplexity models are being deprecated?

Sonar Reasoning (retires 15 Dec 2025).