Waiting for engine...
Skip to main content

API usage limits

REST implementations of the Agentstudio API enforce the following usage limits:

Per-Account Rate Limit​

Each Boomi account gets its own independent bucket. One account's usage does not affect another. These limits are aggregated over all surfaces, that includes the Agentstudio UI, Agent Step and direct API invocations.

API GroupLimit
Design APIs20 requests/second, burst 40
Execute (session) API20 requests/second, burst 40

Account-Level Overrides​

For more information about the rate limits & for custom rate limit requests, contact your account administrator.

Tier-Based Invocation Rate Limit​

Applies only to AI agent session invocations (/execute/api/v1/session).

Account TierLimit
Paid100 requests/minute
Free10 requests/minute

Tier is determined automatically from the account's subscription features.

When exceeded: HTTP 429, Retry-After: 60 seconds.

Token Budget​

Tracks agent input token consumption per account within a rolling 60-second window. Applies only to the agent execute (session) endpoint (/execute/api/v1/session).

Account TierToken Budget per Minute
Paid350,000 tokens
Free50,000 tokens
  • If the budget is exhausted, further session requests are rejected until the window rolls forward.
  • Failed requests (where the LLM was never called) do not count against the budget.
  • When exceeded: HTTP 429, Retry-After: 60 seconds.
On this Page