Skip to main content
Every asset uploaded to Reloops is automatically analyzed to produce Smart Metadata - an AI-generated description and a set of tags that describe what’s actually inside the file. This is what powers semantic AI Search and keeps your DAM organized without anyone manually tagging files.
Smart Metadata panel with AI description suggestion and generated tags including logo, branding, gradient, purple, cyan, infinity symbol, icon, stacked logo, vertical, cropped, black background, vector style, minimal

​
What gets generated

When an asset finishes uploading, Reloops runs it through the AI pipeline and produces:
  • Description suggestion - a short natural-language summary of the asset (composition, subject, style, colors, background, layout).
  • Generated tags - a list of single and multi-word tags covering subject, format, color, style, and context.
  • Captions / OCR - any text visible inside the file is extracted so it becomes searchable.
  • Visual analysis - shape, color palette, and content signals used for semantic matching.
All of this is stored alongside the asset and used by the search index.

​
Reviewing and editing Smart Metadata

  1. Open any asset from a project or from search results.
  2. Expand the Smart Metadata panel in the asset sidebar.
  3. Click the AI description suggestion to accept it as the asset description, or edit it inline before saving.
  4. Generated tags can be removed individually, and you can add your own manual tags at any time.
Manual edits always take precedence over AI suggestions, but the AI-generated values stay in the search index so the asset remains discoverable either way.

​
When tagging runs

AI tagging runs automatically when:
  • A new asset is uploaded to any project
  • A new version is added to an existing asset (the new version is re-analyzed)
  • You manually trigger Regenerate from the Smart Metadata panel

​
AI credits

Smart Metadata generation uses AI credits from your workspace balance. Each asset analyzed (or re-analyzed) consumes credits based on file type and size - videos cost more than still images because every keyframe is examined. You can monitor your credit usage and remaining balance in Workspace settings -> Billing & limits. If you run out of credits, uploads still succeed but new assets won’t be auto-tagged until credits are topped up or the next monthly allowance resets.

​
Why it matters

  • Zero manual work - the team never has to stop and tag files
  • Better search - assets are findable by what they look like, not just what they’re named
  • Consistent metadata - everyone in the workspace gets the same structured tags
  • Faster client delivery - pulling together a deliverable becomes a one-line search instead of a folder hunt