Local · Open source · AI-assisted
Cull, develop and remember your best bird photos. On your own machine.
Point Finchory at a folder of RAW and JPEG frames. It finds the bird, develops the RAW, scores every shot against fixed professional references, names the species when it is sure, and stacks near-identical bursts behind their best frame. Nothing uploads. You make every final call.
Many AI culling and scoring tools send your library to the cloud. Finchory does all of it on your machine.
A desktop app that runs on your own computer. In closed beta now, with a public release on the way.
Runs entirely on your own machine · sample photography shown
Sample photography
Built for photos like these
The bird photography below is Creative Commons sample photography, shown for illustration. It is not the app's output. Finchory works on your own frames: it finds the bird, develops the RAW, scores every shot, and names the species when it is sure, all on your machine.
Sample photography, shown for illustration only. Photo credits in the footer below.
01.What it does
From memory card to keepers
One pass over a folder handles the tedious work; fast keyboard-first views handle the judgment calls.
Cull to the keepers
Near-identical bursts collapse behind their best frame, picked by an image-quality model. A rapid keyboard cull and a side-by-side Compare view make the final pick fast.
One honest score
Every photo earns a 0-100 score on an S-F tier scale, anchored to a fixed set of professional reference images. The same shot scores the same for everyone. It is never tuned to flatter your library.
Species ID that admits doubt
Region-aware identification names the bird only above a confidence floor. Below it, the label is "Unidentified" and the frame is queued for your review. It never guesses to look smart.
Develops the RAW
A warm, punchy grade recovers exposure, white balance, tone and color, with noise reduction that adapts to ISO, so you judge each frame at its real potential, not at its flattest.
eBird-aware
Live eBird data sanity-checks identifications against where you actually shot, flags how notable a species is for that spot, and marks your life-list firsts. Optional eBird features use your own free eBird key.
Remember the good ones
A map of where you shot, a profile with a Hall of Fame of your best frames, and keeper exports: path lists, manifests, eBird-ready CSV, and XMP sidecars for Lightroom, Darktable or Capture One, right from the app's export dialog.
The actual app
What you actually use
These are real Finchory screens, captured on a small sample library so we can show them publicly. The interface, scores, and identifications are exactly what the app produces; the photos are demo data, not a real collection.
Rapid, keyboard-first triage. Keep, skip, or reject one frame at a time, with the rest of the burst a tap away.
Confirm the species. Finchory shows its best guess and confidence, and asks rather than guesses when it is unsure.
Every frame developed, scored, and organized. Filter by species, score tier, location, or date, and sort to your best.
See where each frame was shot. Pins are drawn from your photos' own capture data, all read locally.
02.How it works
Four steps, no cloud
Point it at a folder
RAW and JPEG straight off the card. Indexing runs on your machine; nothing is uploaded anywhere.
It does the heavy lifting
Finds the subject, develops the RAW, scores every frame, identifies what it can, and groups the bursts.
You make the calls
Confirm species and pick keepers in the Review and Cull views, built for the keyboard.
Hand off & relive
Export keepers and XMP to your editor; revisit the map and your Hall of Fame.
03.Local-first, by design
There is no server.
Your photos never leave the folder you point Finchory at, because there is nothing to upload them to. The work happens on your own computer, online or off.
Nothing is uploaded
No account, no telemetry, no cloud sync. There is no Finchory server for your photos to travel to.
Runs fully on your machine
Indexing, developing, scoring, and identification all run locally. An internet connection is optional, used only for eBird checks if you add your own key.
Open source, AGPL-3.0
Finchory is free and open source under the AGPL-3.0 license. No lock-in and no proprietary black box.
You do not have to take our word for it. Watch your own network monitor while Finchory runs and you will see your photos never leave your machine.
Photos uploaded
On your own device
in · XMP sidecars out
Sample photography
Finchory finds, develops, and scores your own frames. Nothing leaves your machine.
04.Why it's built this way
Fair, honest, and yours
A fair score
Photos are graded against a fixed professional anchor set, so the same shot earns the same score for every photographer. No grading on a curve, no inflation as your library grows.
An honest label
A wrong species name is worse than none. Finchory only names what clears its confidence floor and its location sanity check; everything else is flagged for you, plainly, as Unidentified.
Truly local
No account, no subscription, no telemetry, no upload. Your photos and your data stay on your computer, and the whole thing is open source under the AGPL-3.0 license.
Free · Local · Open source
Finchory is in closed beta
It runs entirely on your own computer, points at a folder of bird photos, and does the rest. A public release under the AGPL-3.0 license is on the way, and the download will live here when it is ready.