A folder of raw footage is easy to fill and hard to search. Filenames may tell you when a clip was recorded, but not whether it contains the coastline at sunset, a particular person, or a sign you remember seeing. Finding that moment often means scrubbing through clips or relying on tags someone added earlier.
Reelary is a desktop app for Mac and Windows that aims to make that search happen through descriptions instead. It indexes video and photos on your machine, then lets you look for visual content in ordinary language. The developer says files and the index stay local rather than being uploaded. That makes Reelary a possible fit for people with a large personal or production library who want to find shots without first organizing every file by hand.
From a folder to a search result
The basic workflow starts with a folder or external drive. Reelary says it uses a local vision model to process photos and two-second slices of video, producing descriptions that can be searched by meaning. A request such as a drone shot over a coastline at sunset is meant to return relevant moments, rather than only clips whose filenames contain those words.
That is the practical distinction from a conventional folder browser: you search for what appears in the footage, not just where you put it or what you named it. The product also searches text visible in frames, so a sign, title card, or label can be a search clue. The developer describes indexing and search as available in the free plan, with no account required.
This is still an indexing workflow, not magic retrieval from an unprocessed drive. Reelary needs to inspect the library before its visual search can help. Keeping that work on the computer is useful for privacy and offline access, as the developer claims, but it also means the local machine is part of the process. The supplied information does not specify how long indexing takes or what hardware is needed, so those are worth checking against the size of your library before making it part of a production routine.
Search can lead to a rough cut
Reelary goes a step beyond returning matching clips. Its description says you can turn a script into a rough cut on a simple timeline, then trim, reorder, and caption the selected shots. The result can be exported as an MP4, or handed off as an FCPXML or EDL project for Premiere Pro, Final Cut Pro, or DaVinci Resolve.
That makes the app relevant not only to someone trying to locate one shot, but also to a creator assembling a first pass from existing material. The landing page also describes a local AI voice-over option. These are claims about the product's workflow, not evidence that a generated cut will be ready to publish without editorial review. The timeline is presented as a way to assemble a rough cut, and the handoff formats point to a continuing role for a dedicated editor when a project needs detailed finishing.
What stays local, and what costs extra
The local approach is central to Reelary's pitch: the developer says its models, index, and media remain on the user's machine. The app is available for both macOS and Windows. This will matter to people who cannot or do not want to send source footage to a cloud service, though local processing also ties search to the computer and drives where the library is available.
There is a free plan with a paid upgrade. The free tier is described as allowing indexing and search, while exports are capped at 720p and carry a light watermark. The paid tier adds features including dialogue search, finding people across clips, location search, LAN sharing, and full-resolution or NLE project exports. So the free option can test the core footage search, but it is not equivalent to an unrestricted export workflow.
A fit for searchable archives, not every edit
Reelary's strongest use case is a library where the problem is finding material: a filmmaker's archive, a creator's growing collection of clips, or a photographer's mixed media folders. Descriptive search can reduce the need to remember exact filenames or maintain exhaustive manual tags, while the rough-cut tools offer a starting point after the right material is found. You can learn more about Reelary.
It is less compelling if you need a complete editing environment, or if you expect search to work before the app has indexed your files. The free export restrictions matter too if finished deliverables are the goal. Reelary is for people who want local video search and a quicker route from scattered footage to a shortlist of usable shots. It is not a substitute for a full finishing workflow, nor a way to avoid the initial work of processing a large library.
Search your own footage by describing a shot, fully local