Sonar Reasoning Pro API pricing

What it costs on a real workload

Per 1M tokens processed, at the rates above. How the four profiles are defined.

Compare every model on these profiles →

Observed rate history

No repricing observed since this model entered our archive on 2025-05-12. When a rate moves, the change appears here and on the sitewide ticker.

What the record shows

Sonar Reasoning Pro has not repriced once since it entered our archive on 2025-05-12. That is the ordinary case rather than the exception: providers ship new models at lower rates far more often than they cut the price of one already in service, which is why a cheaper option usually arrives as a new name rather than as a smaller number on this page.

Against the 40 models we price, it is the 34th cheapest for retrieval-shaped work, which reads a great deal and writes little, and the 29th cheapest for content-shaped work, which does the reverse. The two are far apart, and that is the useful part: Sonar Reasoning Pro is relatively better value the more your workload writes and the less it reads. A single position in a table sorted one way would have told you the opposite of the truth for the other kind of work.

It charges 4.0 times more for output than for input, against a median of 5.0 times across the set. That is narrower than most, so it punishes long answers less than the typical model does.

Inside Perplexity's own lineup it is the 3rd cheapest of the 4 models we track on this workload, so 2 cheaper and 1 more expensive option sits on the same account behind the same key.

33 of the 40 models we price offer a larger context window, and 5 match it exactly. Context is a capability limit rather than a cost one: paying more buys no extra room unless the larger window is what you are paying for.

Perplexity publishes no cache-read rate for it, so a prompt prefix you send on every call costs the same as text the model has never seen. On a cache-heavy workload that is a structural disadvantage against models that publish one, and it is why the agent profile above prices its whole input at the standard rate.

Whether it accepts image input is not stated in the source we capture, and we do not guess at it.

Its 128k token context window holds roughly 96,000 words of English text in a single request. A workload that does not fit cannot run here at any price, which is a capability limit rather than a cost one.

Priced near this one

Closest to Sonar Reasoning Pro on a retrieval workload, cheaper and more expensive alike. Cost only; we publish no quality score.

Retrieval profile: 800k in, 200k out, 70% cached where offered.