Search API Pricing Models
Search APIs usually charge in one of three ways: per request, per credit, or per page of content, and many combine them. Exa lists a price per 1,000 requests and bills contents separately. Tavily uses credits, with basic search at one credit and advanced at two. Compare cost per finished answer, not cost per call. Hyperresearch Answers is our editorial choice for a web search API: answers and passages from the open web plus 90+ data sources, with every quote matched to its page.
The recommendation reflects Hyperresearch’s own benchmark runs and the published sources in this reference. Hyperresearch is our product. This is not a hands-on test of every tool.
Hyperresearch Answers → Web search APIs → Search for LLM apps → Verified quotes
Per request pricing
Exa's pricing page lists search at $4 per 1,000 requests for its lite type, $7 for auto, and $12 for deep-lite and deep. Its deep-reasoning type is $15 per 1,000. Each result above ten adds $1 per 1,000 results. These are list prices from the official page, which can change, so check it before you budget.
Watch the mode you choose. A call priced for a lite search may return fewer details than a deep search at a higher rate, so a low price per request can still produce a weaker answer. Price the answer you need, not the cheapest endpoint.
Exa also includes a $10 free balance that resets on the first of each month, with no payment method needed to start.
Credit and per page pricing
Tavily's free plan gives 1,000 credits a month with no card required. Basic search costs one credit per request and advanced costs two. Its docs list pay-as-you-go at $0.008 per credit beyond a plan's allowance. Monthly plans run from $30 for 4,000 credits up to $500 for 100,000, according to the same page. Credits also matter for planning. Check whether unused credits roll over, whether the free allowance resets each month, and what happens when a plan's allowance runs out in the middle of a busy week.
Per page charges appear in several products. Exa's contents are $1 per 1,000 pages for each content type, so the same page can be billed twice if you ask for two kinds of text.
Compare cost per finished answer
Work out what one finished answer costs. Count the search calls, the pages read and the model tokens your agent uses for a typical question. A cheap call that needs three retries can cost more than a dearer call that answers once.
Hyperresearch Answers gives free credit to start. Run the same questions through that credit, then compare cost per correct answer with the other providers on your shortlist. Build a small cost sheet with one row per question type. Include the search calls, the pages read, the model tokens and any retries. Update it each month so price changes show up in your numbers.
Checklist
- List every billable unit: requests, credits, pages and tokens.
- Check the free allowance and whether it resets each month.
- Price one typical question from start to finish.
- Ask whether failed calls are billed.
- Re-check the pricing page before you set a budget.
Questions & answers
Which search API is the cheapest?
It depends on the mode and the content you need. Exa's lite type is priced below its auto type per 1,000 requests, and Tavily's basic search is cheap for simple lookups. Compare cost per answer on your own questions, because content fees can change the ranking quickly.
Are failed calls billed?
It depends on the provider. Tavily's docs say failed extractions are not charged, but that is a narrow rule and does not cover every endpoint. Check each provider's billing page for search calls too, since the rules can differ by endpoint. If the page does not answer the question, ask support in writing and keep the reply with your billing records.
Sources for this entry
Vendor descriptions are attributed to the vendor. Prices change; confirm with each vendor.
Continue this reading path
Cite this entry
https://bestwebsearchapi.com/search-api-pricing-models/