Create each key on its own page, paste it below, then copy the finished lines into your project's .env file. What you paste stays in this tab. It is never sent to Wirestack or saved.
Tavily rate limits are now set per organization, or per user for personal accounts, and every API key under that account shares one limit. Development and production keys draw from the same pool. Before, the limit depended on each key's environment: 100 requests per minute for a development key and 1,000 for a production key.
The limit now follows your plan:
Researcher (free): 100 requests per minute
Pay as you go enabled, Project, Bootstrap, Startup: 1,000
Growth, Scale: 1,500
Enterprise: custom
These stay the same on every plan: /crawl 100, /map 100, /research 20 requests per minute, and the usage endpoint 10 requests per 10 minutes. Going over returns HTTP 429, sometimes with a Retry-After header. Tavily says existing keys keep working and no code changes are needed.
Prerequisites
A Tavily API key in TAVILY_API_KEY
Know which plan your organization is on
Steps
Count every app, test suite and agent that uses a key from your organization. They now share one limit.
If tests or local runs use a development key, keep them from spending the requests your production app needs, for example by throttling test runs.
Handle HTTP 429: wait the number of seconds in Retry-After when it is present, then retry.
If the shared limit is too low, move to a plan with a higher one.
Test it
Send a burst of search requests from two different keys in the same organization. Once the combined count passes your plan's per-minute limit, both keys get HTTP 429.
Not confirmed
The changelog gives only the month (November 2026), not the day. It does not say whether any account's total limit went down.
Wirestack