Will rapid back-to-back searches get throttled?

They can. Tavily Search is documented at roughly two requests per second and returns HTTP 429 on bursts, with exponential backoff as the expected response. A conversation that needs several searches should space them rather than firing them at once.

Does a crawl cost the same as a search against our Tavily credits?

They are metered separately. Get Tavily Usage reports credit consumption, limits, and per-endpoint usage across search, extract, crawl, map, and research, so you can see which operation is consuming the budget rather than only the total.

What happens if the Tavily connection stops authenticating?

Tavily Search returns a 401 when the connection is not active. Lyro surfaces that as a retrieval failure instead of answering from nothing, and hands the conversation to a human agent with the history attached if the customer still needs the answer.

Can we keep Lyro from searching the whole internet?

Yes. Tavily Search supports domain filtering alongside search depth, content type, and result count, and Tavily Crawl adds its own domain filtering with depth and breadth limits, so retrieval can be pinned to your own site and a short list of approved sources.

Tavily MCP integration

Searches the web, parses a linked page, maps or crawls a site, and reports credit consumption endpoint by endpoint.

Connect Tavily See all integrations

5actions available

What your agent can do with Tavily

Every action runs live through MCP. Nothing to build, nothing to maintain.

Tavily search
Lyro performs a targeted web search to find current information customers are asking about. Support agents answer research questions instantly instead of telling customers to search themselves.

See all 5 actions

How businesses use Tavily + Lyro

Each card is one request a support team gets, and the Tavily actions Lyro runs to close it.

Answer from the live web, filtered to sources you trust
Lyro runs a Tavily search with domain filtering and a set search depth, so an answer about something newer than the knowledge base cites pages you allow rather than whatever ranks.
Tavily SearchTavily ExtractGet Tavily Usage
Read the page a customer pasted into chat
Lyro extracts and parses the content behind a URL, with optional image extraction and reranking against the question asked, so the reply responds to the actual page rather than its title.
Tavily ExtractTavily SearchTavily Map Website
Map a documentation site before answering from it
Lyro scans a site and returns a structured list of its URLs without pulling full content, giving a fast picture of what a competitor or a partner's docs actually cover.
Tavily Map WebsiteTavily CrawlTavily Extract
Crawl a domain against a plain-language instruction
With depth and breadth controls and an instruction such as pricing pages only, Lyro traverses a site in parallel and extracts just the pages that match, instead of downloading everything.
Tavily CrawlTavily ExtractTavily Map Website

How it works

Get started in 3 steps

Connect once, then just ask. There is no workflow builder to learn and nothing to maintain — Lyro reads the Tavily actions it has and picks the ones a request needs.

01
Connect Tavily
Authorize the Tavily account your team already uses — one consent screen, no API keys, no mapping tables. Lyro can only do what you granted that account, and you can disconnect it at any time.
02
Tell your agent what you need
Describe the job the way you would hand it to a teammate. Lyro maps it to the Tavily actions that close it and chains as many as the request needs.
03
Watch it work
The agent runs the actions inside the conversation the customer is already in, so nobody copies data between tabs and your team can take over at any point.
Get started free

Tavily · Lyro

Everything else about Tavily

Setup, permissions, and the limits of what Lyro can do inside Tavily.

They can. Tavily Search is documented at roughly two requests per second and returns HTTP 429 on bursts, with exponential backoff as the expected response. A conversation that needs several searches should space them rather than firing them at once.

Every action available in Tavily

All 5 actions your agent can call on Tavily, straight from the live MCP connection.

Tavily crawl
Perform intelligent graph-based website crawling with parallel path exploration and content extraction.
Tavily extract
Extract and parse web page content from specified URLs using Tavily's extract endpoint.
Get tavily usage
Retrieve API key and account usage statistics from Tavily.
Tavily map website
Map a website and discover its pages.
Tavily search
Use this to perform a web search via the Tavily API; offers controls for search depth, content types, result count, and domain filtering.