TokenTally

Perplexity API Platform

Sonar cost planner

Understand how the $5/$8/$12 per-1K request tiers stack on top of Sonar’s token spend.

Last pricing check: Mar 13, 2026

$1.00 per 1M prompt tokens$1.00 per 1M completion tokens128,000 token context

Why teams choose this model

Lightweight sourcing copilots
Automated fact checkers
Consumer news + sports digests
Open main calculator

Scenario planning

Realistic cost examples

Numbers use Sonar pricing

Daily news brief

Marketing teams auto-generate current-event rundowns with low context search for speed.

Per request

$0.001

Per month

$45.00

Tokens sent

45,000,000

480 prompt tokens520 completion tokens45,000 requests/mo

Quick fact concierge

Support reps tap Sonar for grounded answers inside CRM macros.

Per request

$0.0004

Per month

$48.00

Tokens sent

48,000,000

220 prompt tokens180 completion tokens120,000 requests/mo

Streaming summaries

Content apps summarize shows/books with references under the $8 medium-context tier.

Per request

$0.0011

Per month

$40.66

Tokens sent

40,660,000

650 prompt tokens420 completion tokens38,000 requests/mo

Compare with

Perplexity

Perplexity Sonar Pro Pricing Guide

Updated Mar 13, 2026

Anthropic

Claude 3.7 Haiku Cost Calculator

Updated Mar 13, 2026

Anthropic

Claude Haiku 4.5 Pricing Guide

Updated Mar 13, 2026

FAQs

What’s with the per-request fee?

Perplexity bills $5/$8/$12 per 1K requests for low/medium/high search context. TokenTally records those tiers so you can choose which assumption to apply.

Does Sonar support custom tools?

You still pay $0.005 per `web_search` and $0.0005 per `fetch_url` call. Keep prompts tight so the agent doesn’t over-call tools.

How big is the context window?

Sonar exposes 128K tokens, so you can pass transcripts and cached prompts without truncation.

Pricing sources