September brought several big additions to Gumnut, along with more polish throughout the web app. I’ve been continuing to build on feedback from early users, making it easier to get photos into Gumnut, organize them, and do more with them.

You can now create and switch between multiple photo libraries in Gumnut. Keep your personal photos in one library, and create another for a family collection or a shared project. If you manage photos for other people, you can create a library for each client and share it with them.

Gumnut has true library sharing: everyone has access to the same photos, albums, people, and descriptions. When you edit a photo, identify someone in it, or update its description, that work is part of the shared library and is available to everyone with access.

That’s a fundamental difference from Google Photos’ partner sharing. Your partner can save photos into their own account, but later edits don’t update those saved copies, and face labels stay private to each account. In Gumnut, you and your family — or you and your client — work with the same collection and its organization. Sharing access doesn’t create another copy of the library or multiply its storage usage.

You can invite people using a link or QR code. Viewers can browse the library, while Collaborators can contribute photos and help organize it. As the owner, you can manage members and their roles, and set expiration dates or limits on how many people can use an invitation.

Browsing and uploads use the library you’ve selected, so you can keep each collection separate.

September brought support for connecting Grok and Muse, alongside ChatGPT and Claude, plus better compatibility with other MCP clients.

Muse has personally been my favorite new integration. Try some prompts like these, and see how Muse can help you manage your photos in the background!

I also published the Gumnut plugin in the official ChatGPT marketplace, making it easier to connect your photo library. Just install the Gumnut plugin in ChatGPT and follow the prompts.

In August, I added support for non-destructive editing through Immich and the API. In September, editing arrived in the Gumnut web app too.

You can now crop, rotate, and mirror photos directly in your browser, on desktop or mobile. Saved edits show up throughout the app, including the timeline and album covers, without needing a refresh.

Gumnut preserves the original, so you can reopen an edit, make further adjustments, or revert back. You can also choose whether to download the edited photo or the original.

Getting a large collection into Gumnut should be straightforward. In the web app, you can now select a whole folder to upload, with known non-media files skipped automatically. You can also add more files or folders while an upload is running. They’ll appear in an “Up next” queue, where you can remove batches you no longer want to upload.

The web app is just one way to get photos into Gumnut. You can still upload through the API — a common choice for business workflows — or save photos via MCP, for example after an AI agent transforms a photo and saves the result back to your library. You can also sync photos from your phone using the Immich mobile app.

For collections on your Mac or a mounted NAS, there’s also the contributed, open-source GumnutUploader macOS client. It scans the folders you choose, lets you review what will be uploaded, and uploads only files your library doesn’t already have. This contribution isn’t officially maintained by the Gumnut team.

I continued polishing search and browsing, with an always-open filter rail in Search, smoother transitions into and out of the photo viewer, and better timeline navigation. Immich compatibility also moved to version 3.2.0.

For the full list, take a look at the September changelog or the detailed updates.

And of course, I’d love your feedback. Sign up for Gumnut here and let me know what you think!