Activepieces sync webhook 408 at 30 seconds: use Sume async
An Activepieces /sync webhook returns 408 when no response arrives in 30 seconds. Call Sume in async mode, which answers 202 with a job id at once.

Behind an Activepieces /sync webhook, call Sume with the default async mode, not sync. Async returns 202 with the job envelope and a job id straight away, so your Return Response step answers well inside Activepieces' 30 second window and the render finishes on its own.
Activepieces facts are from its limits page; Sume facts from Jobs and results, both read 2026-09-30.
When does Activepieces return 408?
The limits page sets the sync webhook response timeout at 30 s on Cloud. For URLs ending in /sync, a run that fails answers at once with HTTP 500. HTTP 408 is returned only when the run never produced a response within the timeout, which covers a flow still running or one that finishes without a Return Response step.
Why is Sume's sync mode not the fix?
Sume's sync mode waits up to wait_timeout_seconds (max 30) for a terminal state, so on a long render it can use the whole Activepieces window and still return a non-terminal job. The mode never changes how long the job takes to run.
| Sume mode | Returns to your Return Response step | Inside 30 s? |
|---|---|---|
async (default) | 202 and the job id at once | Yes |
sync | Envelope after up to 30 s, maybe not terminal | Not guaranteed |
webhook | 202 and the job id; callback stored | Yes |
What if a sync call returns an unfinished job?
Poll, do not resubmit. A non-terminal sync response still carries the job id, status_url and a sync.timed_out flag; a repeat submit must reuse the same Idempotency-Key.
What does the flow look like?
Trigger on /sync, create the Sume job in async mode, and Return Response with the job id and status_url. The caller, or a later flow, reads the result from there. See the 10-minute run limit for pausing a long run.
Sources
Related posts
More in Developers
- Can an AI agent debug a failed webhook on Sume over MCP?
Over MCP an agent can read job status, events and delivery status to see why a webhook did not arrive. Redeliver is a REST call that needs jobs:write.
- Agent loops on Sume: tool calls spend writes, polls spend reads
An MCP tool call spends the write budget once for the run it creates; a jobs_status poll spends none. Reads have their own, much larger budget.
- AI SDK isLoopFinished vs the 20-step cap for Sume video jobs
ToolLoopAgent stops after 20 steps by default. How many jobs_wait calls a Sume video render needs, and the per-call guards to keep if you lift the cap.
- AI SDK MCP client close() in onEnd: Sume jobs keep running
Closing the AI SDK MCP client in onEnd ends the tool connection, not the Sume jobs already submitted. Track job ids, then read or cancel them.
Written by Sume