Tavily
Use Tavily with AI apps and agents through Vercel Connect. Follow a complete setup guide for secure runtime access.
Install the skill
$ npx skills add vercel/vercel-plugin --skill vercel-connectConnect to
Tavily$ vercel connect create tavily --name acme-tavilyUse its data in agents(opens in new tab) and apps(opens in new tab)
import { getToken } from '@vercel/connect';const token = await getToken('tavily/acme-tavily');
Use Tavily in apps and agents
Search and retrieve information from web sources.
Why use Vercel Connect with Tavily?
Credentials on demand
Request Tavily credentials only when your application or agent needs them instead of copying secrets into application code.
Access for the right subject
Keep app-scoped Tavily access behind a reusable connector controlled by your team.
Centralized credential management
Manage Tavily credentials in one place and retrieve them through a consistent runtime API.
Vercel-native controls
Bind connections to the projects and environments that need them while Vercel OIDC authenticates each runtime request.
Use Tavily with your stack
Follow a complete setup for the Connect SDK or a supported agent framework.
Add Tavily to an existing Vercel application with the Connect SDK.
Install the Connect SDK
npm install @vercel/connectCreate the Tavily connector
Run these commands from the project that will use the connection.
vercel linkvercel connect create tavily --name acme-tavilyvercel env pullRequest Tavily credentials
app/api/route.tsimport { getToken } from '@vercel/connect';const token = await getToken('tavily/acme-tavily');