Skip to content
actionlake
Start free

Connectors · Developer & infrastructure

Tavily MCP server for AI agents.

actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Tavily and the rest of your systems. It gives your agent 10 Tavily actions (8 reads and 2 writes) and holds the Tavily credential itself, so the agent never does. It paces Tavily's limits and records every call.

Connect Tavily freeSee pricing

100,000 free calls a month. No card, no email, no password: press Start a free sandbox on the next screen and keep the key it shows you.

claude mcp add --transport http actionlake https://connect.actionlake.com/mcp
the first call
do {  action: "tavily.logs.list",  intent: "check which Tavily account is connected"}# List usage logs for your API keys · read · 1 upstream call
  • 10actionsin one signed file
  • 8readlook, never change
  • 2writecreate and update
  • 0wait for a yesdestructive, confirmed first

Set up

Connect Tavily in three steps.

  1. 01

    Add actionlake to Claude.

    Run the command once in Claude Code; its consent page can create a free sandbox: 100,000 calls a month, no card. In the Claude app, add a custom connector with the URL below and paste your key on the consent page. Creating a new sandbox from the Claude app is not yet tested by us.

    https://connect.actionlake.com/mcp
  2. 02

    Connect Tavily.

    Bearer token. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead. actionlake stores what you paste write-only, sealed under your tenant's key.

    • api_keyTavily API key (tvly-...): app.tavily.com > API Keys
    account { op: "connections.requirements", service: "tavily" }
# free: exactly what to paste, and what is found for you
  3. 03

    Make the first call. Check before a write.

    Ask in plain words; Claude picks the action. Before anything changes, check validates the call and prices it, for free.

    do    { action: "tavily.logs.list" }# list usage logs for your api keys · read · 1 upstream callcheck { action: "tavily.feedback.create", input: { … } }# free: inputs, permissions, cost and confirmations, nothing sent

Tavily at a glance

Sign-in
Bearer token. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead.
You provide
api_key
Limits we pace to
100 calls a minute, per connection
API covered
Tavily API 1.0.0 (OpenAPI 3.0.3)api.tavily.com
Built from
Tavily's documentation, read 23 October 2026docs.tavily.comdocs.tavily.com
Service file
Version 1.0.0, signed (Ed25519)
Cached reads
7 of 8 reads, for up to 5 minutes; a write clears what it affects
Shared entities
1, so ask can line Tavily up with your other systems
Data
Processed and stored in Sydney, Australia. No US region yet

Capabilities

What an AI agent can do in Tavily.

10 actions in 9 areas. Every one carries its effect, so a key can allow the reads and hold the rest.

  • 8readlook, never change
  • 2writecreate and update
  • 0destructivewait for a confirmation

  1. Research

    2

    1 read · 1 write · 0 destructive

    Get a research task's status and report · Start a research task

  2. Crawl

    1

    1 read · 0 write · 0 destructive

    Crawl a website

  3. Extract

    1

    1 read · 0 write · 0 destructive

    Extract content from URLs

  4. Feedback

    1

    0 read · 1 write · 0 destructive

    Submit feedback on a search

  5. Logs

    1

    1 read · 0 write · 0 destructive

    List usage logs for your API keys

  6. Map

    1

    1 read · 0 write · 0 destructive

    Map a website's URLs

  7. Org usage

    1

    1 read · 0 write · 0 destructive

    Get organization usage across API keys

  8. Search

    1

    1 read · 0 write · 0 destructive

    Search the web

1 more area · 1 actions
  • Usage 1

Top actions

The Tavily API, shaped for agents.

The ids your agent types, what each does, and the one name it shares with other systems. 3 of Tavily's 3 action names are shared with at least one other connector, so an agent learns them once.

ActionWhat it doesEffectSame name in
tavily.mapMap a website's URLsInitiate a web mapping from a base URL.readcontent.site_page.list11 connectors
tavily.logs.listList usage logs for your API keysGet usage logs for your API keys.readplatform.activity_entry.list43 connectors
tavily.searchSearch the webSearch for data based on a query.readplatform.search_result.search18 connectors
tavily.crawlCrawl a websiteInitiate a web crawl from a base URL.readthis connector
tavily.extractExtract content from URLsRetrieve raw web content from specified URLs.readthis connector
tavily.usage.getGet API key and account usageGet API key and account usage details.readthis connector
tavily.research.getGet a research task's status and reportGet research task status and results.readthis connector
tavily.org_usage.getGet organization usage across API keysGet organization usage across all API keys.readthis connector
tavily.feedback.createSubmit feedback on a searchSubmit feedback on a search request or session.writethis connector
tavily.research.createStart a research taskInitiate a research task.writethis connector

Try it

Prompts to paste into Claude.

Each one names the Tavily actions it will call, so you know what it reads and what it changes before you send it.

  1. Readread

    List usage logs for your API keys in Tavily and tell me what stands out.

    callstavily.logs.list

  2. Writewrite

    Submit feedback on a search in Tavily. Show me exactly what you will send before you save it.

    callstavily.feedback.create

Governance for Tavily

Every write can wait for a yes.

None of Tavily's actions is destructive. A key can still be held to reads, capped, and set so a person confirms every write.

A destructive action runs only with confirm: true. Under the human policy the agent cannot confirm for itself: the run stops until a named person approves, in the console or from a signed link in an email. The cap on a key is refused before the call is sent, never after.

a key for this connector
account {  op: "keys.create", name: "tavily-agent",  allowed_actions: ["tavily.*"], cap_calls: 2000,  confirm_policy: "human_for_writes"}
Illustrative. Scoped to Tavily only, capped at 2,000 upstream calls, every write confirmed.

Cost

What Tavily costs on actionlake.

Every Tavily action is one upstream call, the only thing actionlake bills. Searching, check, cost, account and any call Tavily refused with 401, 403, 429 or a 5xx are free. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.

  • 1upstream call per actionall 10 Tavily actions
  • 10actions pageeach page of a long list is one call
  • 7reads can be cacheda cache hit costs nothing
  • 100,000calls a month, freepaid plans from US$49 a month for 250,000

See pricing

Action index

Every Tavily action.

10 actions by area: the id your agent calls, what it does, and its effect.

Research2
  • tavily.research.createStart a research taskwrite
  • tavily.research.getGet a research task's status and reportread
Crawl1
  • tavily.crawlCrawl a websiteread
Extract1
  • tavily.extractExtract content from URLsread
Feedback1
  • tavily.feedback.createSubmit feedback on a searchwrite
Logs1
  • tavily.logs.listList usage logs for your API keysread
Map1
  • tavily.mapMap a website's URLsread
Org usage1
  • tavily.org_usage.getGet organization usage across API keysread
Search1
  • tavily.searchSearch the webread
Usage1
  • tavily.usage.getGet API key and account usageread

Questions

The Tavily MCP server, answered.

Is there a Tavily MCP server for Claude?

Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 10 Tavily actions across research, crawl, extract and feedback, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Tavily.

What do I need to connect Tavily?

Tavily API key (tvly-...): app.tavily.com > API Keys. The free call account { op: "connections.requirements", service: "tavily" } lists exactly what to paste.

Can an AI agent change things in Tavily?

Only if its key allows it. 2 Tavily actions write, and none is destructive. A key can be limited to reads, capped, and set so a person confirms every write.

Will an AI agent use up my Tavily API limit?

actionlake paces every call to the limits in the Tavily service file: 100 calls a minute per connection. Repeated reads are answered from cache (7 Tavily actions can be), which never reaches Tavily.

What does a Tavily action cost?

One upstream call, the only unit actionlake bills. A list that pages bills one call per page (10 Tavily actions page). Cache hits, searches, dry runs and calls Tavily refused with 401, 403, 429 or a 5xx are free. The Free plan includes 100,000 calls a month with no card. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.

Which Tavily API does it cover?

The Tavily API 1.0.0 (OpenAPI 3.0.3) at api.tavily.com. It is built from Tavily's documentation as read on 23 October 2026. 10 actions in 9 areas, each one in the signed service file (version 1.0.0).

Does actionlake keep my Tavily data, and where?

actionlake processes and stores data in Sydney, Australia; there is no US region yet. Reads may be cached for up to 5 minutes, and a write clears the reads it affects. Each run's request and response is kept, sealed, for your plan's trace window: 1 day on Free, 3 days on Starter, 7 days on Pro, 14 days on Team, 30 days on Business and 30 days on Enterprise. Your Tavily credential is sealed under your tenant's key, never logged and never returned. Your traffic is not used to train models.

Which AI clients can use it?

Claude Code, tested by us end to end, and the Claude app as a custom connector, tested by us with an existing key. Other MCP clients that support remote servers over Streamable HTTP (MCP revisions 2025-06-18 and 2025-11-25) and OAuth dynamic client registration, or a bearer header, should connect the same way; we have not tested them. Or plain REST, under the same action names.

More connectors like Tavily

Shares the most action names with

Joins with, on shared entities

Browse

actionlake is independent. It is not made or endorsed by Tavily; the name is the vendor's own. Counts are from the Tavily service file, version 1.0.0.

Give your agent Tavily today.

Connect Tavily freeSee pricing

Start free - 100,000 calls/month, no card. No signup. Hard caps on every key. Cached answers are free.