Skip to the page
Get started free

Perplexity Search API pricing in 2026

$5, fast $1, and the switch.

Perplexity's Search API costs $5.00 per 1,000 successful requests, or $1.00 per 1,000 with Fast Search. A request of up to five queries is one billing unit, and a reply with no results is billed. Agent API web search costs $2.50 per 1,000.

SpecStream is the live web data engine for teams and agents. Its Search API answers Perplexity's POST /search in Perplexity's JSON for $1 per 1,000 searches of up to 10 results, page text included. The switch is one address.

Get an API key

New accounts get $10 of credit, no card, which covers 10,000 searches.

SpecStream and the Perplexity Search API, price by price

SpecStreamPerplexity Search API
Price per 1,000 searches$1 for up to 10 results on the fast tier and on every drop-in address, the Perplexity one included. $5 on the default search$5.00 per 1,000 successful requests. $1.00 per 1,000 with Fast Search, set by search_type: "fast"
What one charge coversOne search of up to 10 results. Each started block of 10 more adds $0.0012 on the fast tierOne successful POST /search. A query array of up to five queries is one billing unit, and uses one rate limit unit per query
Replies that cost nothingA busy call answers 429 with Retry-After and is not charged. Every call, charged or not, is in your request log with its priceInvalid, rate limited and upstream failed requests. A successful reply with no results is billed
Results and page text1 to 20 web results a call, 10 by default. max_tokens_per_page fills each snippet with the page's text at no extra cost1 to 20 for web search and up to 50 for people search, default 10. max_tokens and max_tokens_per_page change what comes back, not the price
Price at volumeA $5,000 search pack holds 1,500,000 default or 7,500,000 fast searches, about $0.67 per 1,000 fast searches, and does not expireUsage tiers from Tier 0 at $0 to Tier 5 at $5,000 or more, based on cumulative purchases
Rate limitWhen search is busy, 429 in Perplexity's own error body with Retry-After, and nothing is charged50 query units per second on POST /search, the same on every usage tier
Free to start$10 of credit on sign-up, no card, which covers 10,000 searchesPrepaid API credits, with auto top-up when the balance runs low
Reading pagesThe Extract API: $0.80 per 1,000 pages, and a page with no text is not chargedfetch_url in the Agent API: $0.0005 per invocation, plus the model's tokens
Every match as rowsDeep Search: every match on the live web as a structured row at 5¢, 15¢ or $1, under a maximum you approve firstpeople_search and finance_search in the Agent API at $5 per 1,000 invocations each, plus tokens

Why Perplexity Search API users move to SpecStream

What 100,000 and 1,000,000 searches cost

At Perplexity's list prices, 100,000 standard requests cost $500 and 100,000 Fast Search requests cost $100. On SpecStream's Perplexity address, 100,000 searches of up to 10 results cost $100, with page text in each snippet at no extra cost.

At 1,000,000 searches a month, Perplexity charges $5,000 standard or $1,000 on Fast Search. A SpecStream search pack of $5,000 holds 7,500,000 fast searches, about $0.67 per 1,000, and the searches left never expire. Committed volume on a signed order goes to $2.00 per 1,000 default searches.

The most accurate web search API on both standard boards

The SpecStream Search API scores 97.9 F-score on all 1,000 official SimpleQA Verified questions (978 correct), and 64.0 item F1 and 43.4 row F1 on all 200 official WideSearch tasks, ahead of Parallel, Exa, Tavily and Perplexity. Scores and method are on the benchmarks page.

You see what every call cost

Perplexity bills each successful request, including one that returns no results. SpecStream logs every call of the Search API, charged or not, in GET /v1/account/requests with its status, latency and price_usd, and each charged reply carries the balance after the call in X-Balance-Cents. A call through the Perplexity address is in the log under that route.

Search, pages and rows on one key

The same key reads pages through the Extract API at $0.80 per 1,000 pages, and returns every match on the live web as rows through Deep Search at 5¢, 15¢ or $1 a row. Agents connect over MCP at https://api.specstream.ai/mcp, with search_web and read_pages on the same balance.

Source URL and read time on every value.

Searches that leave from the EU

A key can send its searches out only from SpecStream's server in Madrid. Turn it on with "eu_egress_only": true when you create the key. It applies on /v1/web-search and on every drop-in address, the Perplexity one included.

Other search APIs priced per 1,000 calls

Exa

Exa charges $7 per 1,000 requests on auto and $4 per 1,000 on lite, with the base price covering the first 10 results. SpecStream answers Exa's own calls at its drop-in address for $1 per 1,000 searches.

Tavily

Tavily bills credits at $0.008 each pay as you go. A basic search costs 1 credit and an advanced search 2, so $8 or $16 per 1,000. SpecStream returns 10 results with page text for $1 per 1,000.

Brave Search API

Brave charges $5 per 1,000 requests and includes $5 in free credits every month, at up to 50 queries per second. SpecStream's fast tier costs $1 per 1,000, and the Extract API reads the pages behind the results.

How SpecStream scores

The same numbers as the benchmarks page, from the official tasks and scorers.

SpecStreamParallelExaTavilyPerplexity
Web researchWideSearch, GPT-5.6 Luna60.945.553.047.947.0
Factual accuracySimpleQA Verified, GPT-5.6 Luna9694919494
Factual accuracySimpleQA Verified, GPT-5.6 Sol9997919295

How to move from the Perplexity Search API to SpecStream

1. Create a key under Agents & API in the workspace. A new account gets $10 of credit, no card, which covers 10,000 searches.

2. Set the base address to https://api.specstream.ai/perplexity where your code has https://api.perplexity.ai, and send the SpecStream key as the Bearer token. The request body and the code that reads results[] stay as they are. In Python:

from perplexity import Perplexity

client = Perplexity(
    api_key=SPECSTREAM_KEY,
    base_url="https://api.specstream.ai/perplexity",  # was https://api.perplexity.ai
)

search = client.search.create(query="Who founded Mistral AI and when?", max_results=10)
for result in search.results:
    print(result.title, result.url, result.date)

3. With curl:

curl https://api.specstream.ai/perplexity/search \
  -H "Authorization: Bearer $SPECSTREAM_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "Who founded Mistral AI and when?", "max_results": 10}'

4. Send one query per call. A query list of two or more gets a 422 that asks for one query per call. Read 402 as an empty balance and 429 as a call to send again after Retry-After.

Every parameter and reply field is on the Perplexity Search API compatible page.

Get an API key

Frequently asked questions

Something we didn't answer? Talk to a human

How much does the Perplexity Search API cost?
$5.00 per 1,000 successful requests, or $1.00 per 1,000 with Fast Search, with no token charges. A successful reply is billed even when it returns no results. SpecStream answers the same POST /search for $1 per 1,000 searches of up to 10 results.
Does Perplexity charge per query or per request?
Per request. A query array of up to five queries is one billing unit, though each query uses one rate limit unit. SpecStream takes one query per call, at $0.001 a search of up to 10 results.
What is Perplexity Fast Search?
A search type on the Search API, set with search_type: "fast", at $1.00 per 1,000 requests in place of $5.00, with the same response format. Every SpecStream drop-in address runs the fast tier by default at $1 per 1,000, with page text included.
What does web search cost in the Perplexity Agent API?
$2.50 per 1,000 web_search calls, or $1.00 per 1,000 with Fast Search, plus the model's tokens. fetch_url costs $0.0005 a call and people_search $5 per 1,000. SpecStream's search_web tool over MCP costs $5 per 1,000 searches of up to 10 results.
What is the Perplexity Search API rate limit?
50 query units per second on POST /search, the same on every usage tier, with a five query array using five units. When SpecStream's search is busy it answers 429 with Retry-After and charges nothing.
Does the perplexityai SDK work with SpecStream?
Yes. Pass base_url="https://api.specstream.ai/perplexity" and a SpecStream key, and client.search.create reads the reply into its own objects. New accounts get $10 of credit, no card.

Sources

SpecStream wrote this page. Each Perplexity, Exa, Tavily and Brave figure comes from that company's own page, read on the day shown.

  1. SpecStream pricingspecstream.ai/pricing, read Nov 12, 2026
  2. SpecStream for agentsspecstream.ai/agents.md, read Nov 12, 2026
  3. SpecStream benchmarksspecstream.ai/benchmarks, read Nov 12, 2026
  4. Perplexity pricingdocs.perplexity.ai/docs/getting-started/pricing, read Nov 12, 2026
  5. Perplexity Fast Searchdocs.perplexity.ai/docs/search/fast-search, read Nov 12, 2026
  6. Perplexity Search API overviewdocs.perplexity.ai/docs/search/overview, read Nov 12, 2026
  7. Perplexity rate limits and usage tiersdocs.perplexity.ai/docs/admin/rate-limits-usage-tiers, read Nov 12, 2026
  8. Perplexity FAQdocs.perplexity.ai/docs/resources/faq, read Nov 12, 2026
  9. Exa pricingexa.ai/docs/admin/pricing, read Nov 12, 2026
  10. Tavily API creditsdocs.tavily.com/documentation/api-credits, read Nov 12, 2026
  11. Brave Search APIbrave.com/search/api/, read Nov 12, 2026