What is Perplexity API?
Developer APIs for ranked web results, cited agent answers from many models, embeddings and probability decisions.
Who is Perplexity API best for?
Perplexity API is best for developers whose apps or agents need fresh web results or cited answers from live search.
Grounding your own agent
The Search API returns ranked web results with extracted page text as JSON, with no model and no token charges, for your own model to read.
Cited answers without running a model
The Agent API runs a model from OpenAI, Anthropic, Google, xAI or Perplexity with web search and other tools, and returns a finished answer with sources.
Search as code
The Python Search SDK exposes search as small building blocks that an agent calls from the code it writes, with concurrency and source tracking built in.
How much does Perplexity API cost?
Perplexity API is pay as you go from credits you buy: the Search API costs $5 per 1,000 requests, and the Agent API bills each model's token rate plus tool fees, with Perplexity's Sonar model at $0.25 per million input tokens and $2.50 per million output tokens.
- Ranked web results with extracted page text
- Fast Search at $1 per 1,000 requests
- Up to five queries in one billed request
- Other models from OpenAI, Anthropic, Google, xAI and more at their own rates
- Web search $2.50 per 1,000 calls, URL fetch $0.0005 a call
- Code sandbox $0.03 per session
- pplx-embed-v1-0.6b, 1,024 dimensions
- pplx-embed-v1-4b at $0.03 per million tokens
- Contextualized models from $0.008 per million tokens
- pplx-decider-flash-v1.2; pplx-decider-v1.2 is $0.02
- Output tokens are free
- Web grounding adds $0.001 per search
There is no subscription and no free tier: each API bills per request or per token from API credits you buy, and rate limits rise with the total you have bought. The Orchestrator API, in private preview, bills each model at its token rate with no request fees. Perplexity is moving Sonar chat completions to the Agent API and recommends the Agent API for new projects. Credits can also be bought through AWS Marketplace, and sales offers custom pricing. Prices are in US dollars. Full pricing on docs.perplexity.ai. We checked these prices on .
How do readers rate Perplexity API?
No one has rated Perplexity API yet. Be the first.
Sign in to rate. One rating per account.
What are the best Perplexity API alternatives?
Three tools on Recaply do Perplexity API's job.
Questions about Perplexity API
What is Perplexity API?
Developer APIs for ranked web results, cited agent answers from many models, embeddings and probability decisions.
Who is Perplexity API for?
Perplexity API is best for developers whose apps or agents need fresh web results or cited answers from live search.
Is Perplexity API free?
No. Perplexity API is a paid tool, from $5 per 1,000 requests.
How much does Perplexity API cost?
Perplexity API's paid plans are Search API at $5 per 1,000 requests, Agent API (Sonar model) at $0.25 / $2.50 per million input / output tokens, Embeddings API at $0.004 per million tokens and Decisions API at $0.01 per million input tokens.
What are the best Perplexity API alternatives?
The closest alternatives on Recaply are Exa, Brave Search API and Parallel Search.
Where can I try Perplexity API?
On its official website, perplexity.ai.
Should I use the Perplexity Search API or the Agent API?
Use the Search API when your own model or agent needs raw web results: it returns ranked results with page text for $5 per 1,000 requests and no token charges. Use the Agent API when you want Perplexity to run the model, search and tools and hand back a finished answer with citations.
What happened to the Perplexity Sonar API?
Perplexity now points Sonar users to the Agent API. Sonar chat completion requests still work but are being turned into Agent API requests, and asynchronous Sonar requests are no longer supported. The Sonar model itself stays available in the Agent API at $0.25 / $2.50 per million input / output tokens.
Claude Platform
Claude Code Agent View
Herdr
Nylas
Cline Kanban
Inbound