Privacy
A default Basic film talks only to your Immich server. The app sends no analytics. Hugging Face Hub telemetry and ONNX Runtime's Microsoft telemetry (which sends HTTP by default) are switched off before anything imports them, in the app and in the inference service. The update check stays off until you ask for it.
Model files and the music library come down once, during setup. A few explicit download switches and the optional music backends can also fetch weights on first use, including during a run.
This documentation website counts visits with the maintainer's self-hosted Plausible. GitHub Pages serves the tracking script alongside the docs; visit and click events go to the maintainer's collector. The tracker uses no cookies. The website privacy page lists what it records and has an opt-out button. The app you install sends no analytics.
Optional features can contact other services. You choose which ones, and whether they run on your own hardware or outside your network.
Before adding a service
| Feature | What it receives |
|---|---|
| Text reader | Candidate annotations, dates, places, people/relationship and album names, plus your memory brief. On Full, the period's account and one card per story in the cut (what it was, first names, what its shots show people doing) for the film summary. A special day's title prompt (a trip's too with --llm-title); never a title for a year, month, season, person, couple, birthday, holiday or on-this-day film |
| Caption server | Small picture tiles and video-frame strips |
| LLM captions | Pictures, only with explicit caption_provider: llm |
| Inference service | Previews and sampled frames for classification; generated WAV music for stem separation |
| Render worker | Cut/timing metadata, primary and selected partner API keys, names, titles, home coordinates and network settings |
| Geocoding/maps | Rounded coordinates or map tile requests |
| Generated music | Mood/genre/tempo text; stem separation can send audio |
| Notifications | Run outcome/details, optionally a thumbnail |
| OIDC | The usual login flow to your provider |
| Update check | One anonymous request for GitHub's public release list: no version, install id or setting |
| Model, music and font downloads | Requests for pinned files from Hugging Face and GitHub; no library data |
The reader is disabled by default. Enable it with llm.enabled: true, blank base_url and openai-compatible or ollama for an owned local reader;
a remote URL or hosted provider preset sends text to that endpoint. Configuring a reader does
not enable image captions. Check the full request inventory for exact
hosts, switches and defaults. preflight names enabled geocoding/map destinations; that list
does not cover every outside endpoint. Its service checks also send authenticated probes.
What Immich sees
The app reads metadata, previews and source media. It never changes your originals. With upload off (the default), it does not write to the library. With delivery enabled:
- Upload the finished film as a new asset.
- Tag it
immich-memories/generated, so it is not selected as source footage later, plusimmich-memories/version/<release>andimmich-memories/generated-on/<YYYY-MM-DD>: the release that made the film and the day it finished, on the server's clock. A film's dates end on its last picture, so these say when it was made. A run older than 1.0.0 has no recorded release and gets no version tag. - Describe it: the description is the film's title and subtitle, for example
July 2025. Withtitle_screens.llm_titles: true(experimental) its story summary follows. Withupload.film_tags: true, also tag itimmich-memories/type/<film type>,immich-memories/year/<year>for each year its stories in the cut cover andimmich-memories/story/<name>for each birthday, visit or trip it shows. People get no tag; Immich's faces already say who is in a film. Only new uploads and re-renders get these. An API key withoutasset.updateor the tag permissions still uploads the film, with a warning that the label didn't land. - Add it to the configured album (create the album if needed).
- Move a recognised previous render of the same recipe to trash, if the API key permits it.
immich-memories config test only checks authentication/API compatibility.
Geocoding and maps
Both are off by default:
network:
geocoding: false
geocoding_url: ""
map_tiles: false
Geocoding gets city, village and trip names and names in the film's language. It sends
coordinates rounded to two decimals (about a kilometre) to Nominatim, once per place. That can
include home. Answers stay in the store. Set geocoding_url for your own Nominatim.
Place names are requested in the film's language; a place without a name in that language keeps its local name instead of an English fallback.
Map tiles come from ArcGIS World Imagery for trip maps and location-card backgrounds. Requests reveal the area, including home base. With tiles off, ordinary title/location cards still work. Maps and titles explains the result.
Update check
Nothing asks for a newer release until you do. Check for updates in Settings and
immich-memories version --check ask api.github.com once, when you click or run them. Set
updates.check_daily: true to have the app ask at most once a day and show a banner when a release
is out. The request is an anonymous GET of the public release list; the app compares versions
locally.
Fonts, models and music
Docker includes script fonts. Python installs include the core fonts; other alphabets use one explicit download:
immich-memories titles fonts --install
Font files are pinned and digest-checked. A render never fetches a font. Detector downloads are
off unless allowed; the Compose inference overlay allows them, and the GPU overlay's one-shot
immich-memories-caption-models container fetches the caption model's weights at compose up,
before the captioner starts. models fetch and the web UI's Download models button run the same
download on demand, and fetch the 118 MB music library from the project's GitHub releases
(digest-checked, skipped when audio.no_music is on; the Docker image already has it). A film
never downloads music. ACE-Step and local Demucs can fetch their own weights on first use.
Those downloads contain no library data. Local ACE-Step uses app-pinned immutable Hugging Face
snapshots and refuses incomplete downloads; checkpoint revisions
list the pins. API mode uses the remote music server's checkpoints.
Thumbnails in the web UI
Your browser asks this app for pictures/video, behind the same login. The app fetches or serves cached media from Immich. The API key stays on the server. It only serves a picture one of the configured accounts can read: any other ID gets a 404, even when it sits in the cache.
Cached suggestions and trip summaries are private library data. Their JSON files are readable only by the account running the app, including existing answers when the app reads them. Job progress files can include the finished film’s path. New files and replacements for older progress files are readable only by that account. Opening the web job runner also restricts its job directory, protecting older progress files without changing the shared cache folder above it.
Privacy mode
For sharing a demonstration without showing the actual film:
immich-memories generate --privacy-mode --year 2024
The encoded film blurs pictures, scrambles speech, substitutes names and moves places to a fake city. This changes the film, not network requests. Trip geocoding can happen before names are substituted. The output filename can still contain real information: rename it before sharing.
For a screen share, enable server.enable_demo_mode and turn on the UI's Demo mode eye button.
Each browser has its own choice. This only blurs what the UI displays.
Diagnostic reports
immich-memories report creates a local report, with configured credentials, known personal
names/places, coordinates, addresses, URLs and absolute paths removed. Asset IDs become report-local
hashes. It contains no pictures and sends nothing automatically.
Read it before sharing; automatic redaction is not a substitute for checking the file.
Personal documents
Every tier checks for screenshots and photographed documents, including ID cards, bank cards, letters and forms. The OCR check uses your configured Immich server. Suspected documents are held back, but a pin can override the personal-document check. Review the cut before sharing: a document can escape detection. See picking a shot for the checks.
Everything that can leave, and when
The network request inventory keeps the complete egress table and image-provider details together for operators.
The one picture seat
The ordinary reader receives text. Picture tiles go to the caption server, or to the configured LLM only after an explicit image-caption opt-in. See the inventory.
Files on disk, and cancellation
Technical privacy details cover local file permissions and cancellation. Keep the store backed up.
Fetch once, then block outbound
The offline Basic recipe seeds a model volume before restricting the app to Immich. It includes Docker's same-host internal bridge and a destination-scoped Kubernetes policy, with their routing limits.
Before using the Basic offline recipe
Basic omits the extra GPU/Full detectors and Laya sharing pre-screen. Choosing Family instead of Just us does not add that missing coverage. Review the film before sharing; the audience guide explains the limits.
For an additional text service, see Basic with a local text model.