Ecommerce generation queue

Monitoring background AI generation jobs, checking status and errors, retrying failed jobs, and handling insufficient credits.

What is the generation queue

Every time you launch an AI content generation from Ecommerce (descriptions, images or attributes), the work runs in the background as a job. The generation queue is where you track those jobs: whether they're running, completed, or failed, and where they came from.

What you can do

See the status of every generation job: in progress, completed, or failed.
Filter jobs by their source (for example, which import or generation batch they belong to).
Browse pages of jobs if you have many.
Open the details of a failed job to see the exact error.
Retry a generation that failed.

Checking the queue

From Ecommerce, open the generation queue (reachable from the product table, near the generation actions).
Review the list of jobs and their status.
Use the source filter to narrow the list down, for example to a specific CSV import or a specific generation batch.
Move between pages if the list is long.

Understanding job status

Status	Meaning
In progress	The job is currently being processed; content isn't ready yet.
Completed	The job finished successfully; generated content is available on the product(s).
Failed	Something stopped the job from completing; open the details to see why.

Investigating and retrying a failed job

Find the failed job in the queue.
Open its details to read the specific error message.
If the cause is fixable (for example, a temporary issue), retry the generation from the same screen.
If the cause is insufficient credits, top up your balance before retrying (see below).

Credits and errors

Generation jobs use the same credit rules as the rest of Ecommerce generation (5 credits per description, 5 per image, 1 per attribute completion). If a job cannot proceed because your balance is too low, you'll see a dedicated "insufficient credits" message, and any credits reserved for that job are released back to your balance rather than being lost.

Reserving credits before running a job means retrying a job you've already paid the reservation for won't double-charge you; the system is designed to be safe to retry.

Tips

Check the queue after launching a large batch generation, especially if you selected many products or images at once.
If a job fails due to an external research service being unavailable (used to enrich descriptions with up-to-date information), you can usually retry later once that service is available again.
Use the source filter to keep track of separate batches, especially if you run generations from different imports or templates.

Common problems

Problem	What to check
A job stays "in progress" for a long time	Large batches (many products or images) naturally take longer; check back later or filter to see if other jobs from the same batch have completed.
Job failed with "insufficient credits"	Check your AI credit balance and top up if needed, then retry the job.
Job failed for an unrelated reason	Open the job details for the specific error message before retrying; if it recurs, contact support.
Can't find a specific job	Use the source filter and browse additional pages if your catalog has many recent generations.