Price terms as published

Each line is the vendor’s own wording, split from the price capture checked on 2026-10-14. Nothing here is rounded, converted or re-worded — the full sentences stay in the review below. Lines that carry only a price continue the plan named on the line above, in the vendor’s own order.

3 price terms for Keenable
#Plan / termChecked
1Free: 100K requests/mo2026-10-14
2Pay-as-you-go $4/1K requests2026-10-14
3$1/1K requests at 100+ RPS2026-10-14

One-liner: Keenable is a web search API built for AI agents. You give your agent a search endpoint, Keenable returns structured results. The free tier is 100,000 requests per month; paid usage runs $4 per 1,000 requests.

Verified facts (checked 2026-09-07; pricing re-read 2026-10-14)

Attribute Value Source
Free tier 100K requests/month, granted on signup Official pricing page, checked same day
Pay-as-you-go $4 / 1,000 requests, cloud-only access Official pricing page, same day
High-volume tier $1 / 1,000 requests at 100+ RPS, cloud & on-premises Official pricing page, same day
Funding $26M raised, led by Accel and Conviction TechCrunch, 2026-08-25
SDK ecosystem Official LangChain / LlamaIndex integrations (keenableai GitHub org, MIT) GitHub org, checked by collaborator 2026-09-06
Products Search API + “Time Machine” (historical web index) Official site nav

This is an official-information entry — we have not run production traffic through the API, so no score, and every cost claim below comes from the pricing page as re-read on 2026-10-14, where the 100K free monthly requests, the $4 per 1,000 requests and the $1 per 1,000 at 100+ RPS were all unchanged from 2026-09-07.

The pricing structure is the interesting part

Most search APIs for AI (the older generation) sell monthly subscription tiers: X queries per month, capped. Keenable sells raw metered throughput with a volume ladder:

  • $4 per 1,000 requests is the standard pay-as-you-go rate, cloud-only.
  • $1 per 1,000 requests appears only at 100+ requests-per-second commitment, with dedicated capacity and an on-premises option.

That’s a 4× price drop purely for becoming a heavy user — a sign that search infrastructure for agents is being priced like inference, not like SaaS. The $1–4 per 1K request band is a useful market anchor for 2026 agent infrastructure generally.

What the free tier actually requires

The pricing page’s call-to-action is “Get Free 100K requests” linking to the app login/signup console. The page itself says “100K free requests a month” with no card requirement mentioned — but no-card is not explicitly stated on the pricing page either. Treat “no credit card needed” as unconfirmed until you go through signup yourself.

Both free and $4/1K tiers are cloud-only. On-premises deployment only appears at the 100+ RPS tier, which puts private deployment out of reach for small users.

Who it’s for

  • Agent builders who need live web access inside a LangChain/LlamaIndex/Langflow pipeline and don’t want a per-query subscription
  • AI labs and inference platforms — the 100+ RPS tier at $1/1K exists specifically for them, with on-prem
  • Not a consumer search product: there’s no chat UI to browse, it’s an API you wire into software

Caveats

  • Funding announcements (TechCrunch) describe an indexing ambition — “indexing the web for AI agents” — but Keenable is young; independence from the Google/Bing index underneath is a marketing claim we cannot verify from outside.
  • The free 100K is requests, not tokens or results-per-request; agents that fan out multiple queries per user turn will burn the quota faster than it sounds.

Sources