> ## Documentation Index
> Fetch the complete documentation index at: https://docs.gumnut.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Uploads and processing

> Add files, folders, or your phone's photos and understand what happens after upload

You can add photos and videos through the web app, an Immich client, the API,
or an AI assistant. Select the destination [library](/guides/product/libraries)
first; you need to own it or have the Collaborator role to upload.

## Upload in your browser

Use **Upload** in the [web app](https://app.gumnut.ai) to choose files, or
**Upload folder** to choose a folder. Known non-media files are skipped.
You can add more files or folders while an upload runs; later batches appear
under **Up next**, where you can remove a batch before it starts.

[Exact duplicate uploads](/guides/product/duplicate-detection) reuse the
existing asset in that library. They do not create a second stored copy.

<Frame caption="Upload queue example: the current batch runs while another waits under Up next.">
  <img src="https://mintcdn.com/gumnutai/wUJYRczAyiokIPpS/images/product/upload-queue.jpg?fit=max&auto=format&n=wUJYRczAyiokIPpS&q=85&s=dee90160d21064b7eafe85c0b15ed4e7" alt="Illustrative upload card showing photo thumbnails, progress, and an Up next batch" width="1272" height="1042" data-path="images/product/upload-queue.jpg" />
</Frame>

## Bring photos from other tools

* Use the [Immich mobile app](/guides/immich/overview) for phone backup.
* Use the [REST API](/guides/apis/overview) or
  [SDKs](/guides/sdks/overview) for scripts, imports, and business workflows.
* An [AI assistant](/guides/product/ai-assistants) can save images it edits or
  generates back to your library through MCP.

## What happens after upload

Processing continues in the background. Gumnut prepares thumbnails and reads
metadata such as capture time, camera details, and location when available.
For images, it also detects faces, creates search embeddings, and generates
descriptions of visible content. These results can take time to appear.

Videos support thumbnails, metadata, and playback of the original file;
automatic face, embedding, and description analysis is image-only today.
See [People](/guides/product/people), [Search](/guides/product/search-and-filtering),
and [Map view](/guides/product/map) for ways to explore the results.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.