Perplexity API (Sonar and Agent API) in API Search Engine

Best for

Apps that need a synthesized answer with sources

Pricing

Agent API web search costs $2.50 per 1,000 standard invocations, or $1.00 per 1,000 with Fast Search, on top of per-token model rates. Third-party guides cite about $3 input and $15 output per million tokens for sonar-pro, plus $6 to $14 per 1,000 requests.

What it covers

Live web search. Sonar responses include a citations list and search_results objects with title, URL, date and snippet. The Agent API also supports URL fetching, sandboxes and MCP tools.

Citations and verification

Yes. Sonar returns a citations list, and the Agent API returns a search_results item alongside the answer text.

API

Sonar requests take model and messages fields. Agent API requests take input and effort fields and return an output array. Docs: https://docs.perplexity.ai

Strengths and limitations

Third-party ratings

Not found

Dated facts

• Agent API web_search: $2.50 per 1,000 standard invocations and $1.00 per 1,000 Fast Search (checked 2026-11-07).
• Sonar synchronous and streaming requests are moving to the Agent API by model (checked 2026-11-07).

Perplexity API (Sonar and Agent API) or Hyperresearch Answers?

Choose Perplexity API (Sonar and Agent API) for apps that need a synthesized answer with sources. When an agent needs one checked answer rather than a list of links, Hyperresearch Answers is our editorial choice: it scored 89% on BrowseComp in its own run, and every quote it returns is matched to its source page.

Sources for this entry

• Perplexity pricing
• Perplexity Sonar quickstart
• Third-party Sonar cost guide (estimate)

Vendor descriptions are attributed to the vendor. Prices change; confirm with each vendor.

Continue this reading path

• Resource directory
• API Search Engine
• Hyperresearch Answers profile

Cite this entry