Why is my document still processing?

Every document goes through a pipeline: queued → extracting → chunking → embedding → indexing → done. Memory extraction ("dreaming") then catches up in the background.

What's normal

• Most documents finish in seconds to a few minutes. Large PDFs, videos, and big connector imports take longer.
• Content is searchable as soon as indexing completes — background memory extraction typically catches up within about 15 minutes.
• During heavy load, documents can sit in queued for a while. The queue is automatically swept every few hours, so briefly stuck items clear themselves.

When to worry

Contact support if any of these persist for more than a few hours:

• A document stuck in queued, chunking, or embedding.
• A document showing done with chunks present but 0 memories long after upload.
• A failed status, or uploads that disappear without appearing in your dashboard.

What to include in your report

So we can fix it in one round-trip, include:

• Your organization ID (Console → Settings)
• One or two affected document IDs (and customId if you set one)
• The container tag you uploaded into
• Roughly when you uploaded

Live platform status is at status.supermemory.ai.