Kagi pricing: web search API
Is Kagi yours? Write to us about this page. We invite selected tool providers to our WhatsApp group.
17% complete: 1 of the 6 key fields we checkSee the 6 fields
$12.00per 1,000 searchesWe haven't run Kagi through our web search API benchmark yet.Half of the 569 plans we rank by price cost $3.992 per 1,000 searches or less. This plan costs less per 1,000 searches than 19% of the 569 plans.browserfoundry by keirolabs's $99.99 monthly plan is the cheapest paid plan per 1,000 searches: $0.033.browserfoundry by keirolabs, Context.dev and Exa are the best web search API providers in the Datacircle benchmark.
Plan: Search one of Kagi's 2 plans
Call:
POST / Kagi's only callAPI: Web search API Kagi's only API
About this item
- Live. Kagi searches the web when you call the API.
- 1 of the 6 key fields we check. Text.
- Nothing to buy first. You pay at a threshold.
- One call. You send
querytoPOST /. The call returns the matching web pages.search - Who can get it: with conditions. Keep data: with conditions. Use data: with conditions.
Kagi: web search API alternatives
Kagi alternatives: every web search API provider ranked
We list other providers' plans closest to this plan in price per 1,000 searches. Each plan in the table gives live data.
| This planKagi, pay as you go | Scrapio, $79 monthly | ScrapeGraphAI, $20 monthly | Tako, pay as you go, Pay per request (x402), POST /v1/search, Search, deep | AgentFetch, $80 pack | |
|---|---|---|---|---|---|
| $ per 1,000 searches | $12.00 | $11.85 | $12.00 | $12.00 | $12.00 |
| Completeness | 17% | 17% | 17% | 33% | 17% |
| Billing | pay as you go | monthly | monthly | pay as you go | pay first |
| Pay | at a threshold | $79 | $20 | per call | $80 |
| Enough for | as many as you pay for | about 6,666 calls a month | about 1,666 calls a month | as many as you pay for | about 6,666 calls |
| Freshness | live | live | live | live | live |
| Who can get it | anyone with a Kagi account who adds funds to its API balance (its API setup's "Add funds to your API balance."), on a Family or Team plan only the account owner can open the API portal | anyone at least 18 years old and legally capable of entering a binding contract | anyone who can form a binding contract and whose use is lawful where they and their users operate (its terms: "You may use ScrapeGraphAI only if you can form a binding contract and if your use is lawful in the places where you and your users operate.") | anyone who pays the request, with no account or API key | anyone |
Sources
From each provider's own pricing page and docs, and from what providers tell us. We read this row from these pages on 2026-10-28.
Product details
The key fields this plan's answer can hold, then each filled column in its row on the sheet.
- Date, not in the answer
- Author, not in the answer
- Text
- Summary, not in the answer
- Links, not in the answer
- Answer, not in the answer
| $ per 1,000 searches | $12.00 |
|---|---|
| Per | call |
| Billing | pay as you go |
| Aggregators | none |
|---|
| Also pays for | nothing else |
|---|
| Their pitch | The highest quality search results. Your system is only as good as its data source. Kagi APIs give your system the edge it needs to perform at its best. |
|---|
| Freshness | live |
|---|---|
| Keep data | only the transient caching needed to operate your application: "Store, cache, or build a database of Results beyond the transient caching needed to operate your application. You may not create a persistent copy of, or a derivative index from, the Kagi search index or other Results, and you may not bulk download, scrape, or archive Results." |
| Use data | Within your plan limits, you get a limited, non-exclusive, non-transferable, non-sublicensable, revocable license to call the API and use the outputs it returns, including search results, translations, and summaries ("Results") in your own applications and services, including sending Results to an AI model to answer your users' queries. [...] You may not: Store, cache, or build a database of Results beyond the transient caching needed to operate your application. You may not create a persistent copy of, or a derivative index from, the Kagi search index or other Results, and you may not bulk download, scrape, or archive Results. Use Results to train or fine-tune a search index, ranking system, translation or summarization model, or other service that competes with Kagi, or systematically harvest Results to assemble a training or evaluation dataset. Resell, sublicense, proxy, redistribute, or publish the API or raw Results as a standalone product, bulk feed, dataset, or substitute for direct Kagi API access, unless we agree in writing. Circumvent rate limits, quotas, or access controls, or attempt to access the API by any means other than the documented interface. Reverse engineer the API or use it to probe, scan, or test the vulnerability of Kagi's systems. Use the API for anything unlawful, or in a way that infringes others' rights or violates these Terms. |
| Pay | at a threshold |
|---|
| Batch | 1 |
|---|
| Answer | in the call |
|---|
| You send | query. Optional: workflow, format, lens_id, lens.sites_included, lens.sites_excluded, lens.keywords_included, lens.keywords_excluded, lens.file_type, lens.time_after, lens.time_before, lens.time_relative, lens.search_region, timeout, page, limit, filters.region, filters.after, filters.before, extract.count, extract.timeout, safe_search, personalizations.domains[].domain, personalizations.domains[].kind, personalizations.regexes[].regex, personalizations.regexes[].replacement |
|---|---|
| Method | POST |
| Path | /search |
| Plan | Search |
|---|---|
| Who can get it | anyone with a Kagi account who adds funds to its API balance (its API setup's "Add funds to your API balance."), on a Family or Team plan only the account owner can open the API portal |
The 29 fields the plan's answer can hold
trace ID, node, search time, web results, image results, video results, podcast results, podcast creator results, news results, adjacent questions, direct answers, interesting news, interesting finds, infobox, code results, package tracking, public records, weather, related searches, listicles, web archive results, URL, page title, snippet, date, image URL, image width, image height, result properties