Linkup vs Tavily Pricing per 1,000 Searches

Linkup and Tavily sell the same thing to the same buyer: a web search API that an AI agent can call and get back sources it can quote. Both plug into LangChain, LlamaIndex and the usual agent frameworks, both return content rather than ten blue links, and both publish their prices. The difference is how they charge. Linkup prices each call in dollars against a balance you top up. Tavily sells monthly credit plans, and each endpoint draws a different number of credits.

That makes them hard to compare from the headline. This post lists both price lists as each vendor publishes them, read on 15 October 2026 from Tavily's pricing page and credits documentation and from Linkup's pricing page and docs, then prices five agent workloads on each so you can see where the gap is real and where it disappears.

Linkup vs Tavily pricing at a glance

Tavily's plan ladder is Project at $30 for 4,000 credits, Bootstrap at $100 for 15,000, Startup at $220 for 38,000 and Growth at $500 for 100,000, with Enterprise quoted. Read as a price per credit that runs from $0.0075 down to $0.005, so a basic search costs between $7.50 and $5 per 1,000 on a plan and $8 on pay as you go. Linkup has no ladder. Its only volume lever is a top-up bonus of 10 percent at $1,000, 15 percent at $5,000 and 20 percent at $10,000.

Is Linkup cheaper than Tavily?

For standard search, yes. Linkup's $5 per 1,000 matches the cheapest rate Tavily offers, and Tavily only reaches it on the $500 Growth plan. Below that volume Tavily costs $5.80 to $8 per 1,000 basic searches. Linkup also gives more away: $20 a month covers 4,000 standard searches, against 1,000 basic searches on Tavily's free credits. Tavily can come out cheaper on research calls that stay near its lower bounds.

Five agent workloads priced on both

Each workload below uses the cheapest way to buy it on each vendor at list price, before any enterprise discount. For Tavily that is usually the largest plan that fits plus pay as you go for the rest. Linkup figures include the $20 monthly credit where it matters.

The first two rows are the clean comparison, and Linkup wins both. A production agent doing 50,000 ordinary searches a month costs about $66 less on Linkup, or roughly a fifth, and the prototype barely costs anything while the free credit lasts.

The third row looks like a Tavily win by a factor of five, and it needs a warning. The two "deeper" settings are not the same product. Tavily advanced search is a single search that retrieves more relevant content per result. Linkup deep search runs an agentic, multi-step search loop for questions that need several hops. A team that needs Tavily advanced quality may find Linkup standard is enough, and a team that needs Linkup deep may find Tavily advanced is not. Run twenty of your own queries through both before you trust this row in either direction.

What does a Tavily research call cost compared with Linkup Research?

A Tavily research call costs between 4 and 110 credits on the mini model and between 15 and 250 on pro, which is $0.03 to $2.00 a call at pay-as-you-go rates. A Linkup Research run costs a fixed $0.25, $0.50, $1.50 or $2.50, depending on the size you request. Tavily can be cheaper per report, but Linkup is the one you can put in a budget.

That band matters more than the averages. A thousand pro research calls on Tavily could land anywhere from $120 to $2,000, and you only learn where after the month closes. If research reports are a feature you sell to your own customers at a fixed price, Linkup's fixed run price protects your margin. If research is an internal tool and you watch the bill, Tavily's lower bounds are worth testing. And if the questions are legal ones, about statutes, filings or court opinions, a general web search agent is the wrong instrument altogether, and a dedicated AI legal research tool that searches the case law itself will give cited answers a web index cannot.

Which is better for AI agents, Linkup or Tavily?

Linkup is the better buy for agents that run many standard searches and need a sourced answer on each, because it is cheaper per call and its prices do not move with plan size. Tavily is the better buy for teams that want a monthly plan on the invoice, a map and crawl endpoint alongside search, and a long list of ready-made framework integrations. Both return content an agent can cite.

Two operational details belong in the decision as well. Linkup applies a limit of 10 queries per second per organization, shared by search and fetch, so a fleet of parallel agents should check it before launch. Tavily charges extract and map only for successful URLs, and Linkup does not charge for failed calls at all, including searches that find nothing relevant, which is generous for agents that ask a lot of speculative questions.

Linkup vs Tavily for reading pages you already know

Both vendors will read a known URL for you, and both are the expensive way to do it at volume. Linkup Fetch costs $1 per 1,000 pages without JavaScript and $5 with rendering. Tavily Extract costs $1.60 per 1,000 basic pages on pay as you go, and $1 at the Growth rate. Tavily's crawl endpoint is cheaper still for whole sites, 3 credits per 10 pages at basic depth, but it still bills every page through the same credit pool your searches need.

This is where a search API stops being the right meter. If your agent spends most of its budget reading documentation sites, product catalogs or competitor pages it already knows about, there is nothing to search for. A crawl plan that bills a flat price per rendered page does that job for less. On ClawEngine the Startup plan is $99 a month for about 250,000 pages with JavaScript rendering included, which is about $0.40 per 1,000 against $5 on Linkup Fetch with rendering. Our web crawler API walks a site from a start URL, and the web data API for AI agents page shows how teams split the job: a search API to find the pages, a crawler to read them.

To be plain about the limits: ClawEngine has no search product and no semantic index. It cannot replace either vendor for the "find me sources on this question" half of an agent. It replaces the reading half.

How to choose between Linkup and Tavily

• A prototype or a small internal agent: Linkup. The $20 monthly credit covers 4,000 searches and you never sign up for a plan.
• A production agent with tens of thousands of standard searches: Linkup, by about a fifth at 50,000 searches a month, and by more if you stay below Tavily plan sizes.
• Research reports sold at a fixed price: Linkup, because the cost per run is fixed in advance.
• A monthly plan for procurement, plus map and crawl in the same account: Tavily.
• Heavy use of a deeper search mode: test both on your own queries first, since the two modes are not equivalent.
• Bulk reading of known sites: neither. Use a crawl plan billed per page and keep the search API for finding sources.

For the full price list on each side, see our Tavily pricing breakdown and the Linkup API pricing guide. If Exa is also on the shortlist, Exa vs Tavily pricing prices that pair the same way, and Tavily alternatives covers the rest of the field.