Frequently asked questions

It is a Windows desktop application (with companion support where the line ships it) for local-first analysis of images, video, audio and metadata. It produces structured, open .vtag sidecar files so you can make archives more searchable and connect to tools like catalog software.

An AI file search tool helps you find files using model-derived understanding—scene descriptions, speech transcripts, structured fields—in addition to filenames and folders. VisionaryAI Suite implements that idea as a local-first Windows desktop workflow with open .vtag sidecars where your build supports them. More context: Search files with AI.

Yes, in a workflow sense: analysis on your machine can yield searchable text and fields tied to visual and temporal content. Exact models, codecs and limits depend on your build and licence—validate on your own samples via the trial.

The design centre is local-first analysis on hardware you control. Some network use may still happen (licensing, updates, optional services). Your installer’s privacy text is the binding reference—not this marketing FAQ.

That is the practical aim: AI-generated metadata and semantic memory reduce the need to hand-tag everything up front. You can still review and edit what gets stored.

Semantic media search is querying by meaning and context—what appears, is said or is implied—rather than only rigid keywords or filenames. See the AI file search overview for how VisionaryAI Suite approaches it.

.vtag files are open JSON sidecars beside media. They store structured AI output and provenance so collections stay searchable and portable. Technical detail: .vtag format page.

When transcription is part of your pipeline, speech-to-text can feed search and review together with visual metadata—depending on models and settings in your build.

It is a Windows desktop application aimed at media-heavy collections—not a generic replacement for the OS filename indexer everywhere. The focus is local AI analysis, semantic memory and interoperable metadata.

The design centre is analysis on your own machines where your build is installed. Platform support is expanding—treat the current product sheet as the source of truth for which OS and companion apps apply to your license.

In the local-first workflow, your source files are processed where you run the suite, and results can be written as sidecar metadata on disk. Any optional online feature (licence checks, model downloads, crash reporting) is version-specific: read the privacy section of your build’s documentation; we do not promise a zero-network binary on this marketing site.

A .vtag is an open sidecar file—usually JSON—saved next to a media file. It contains structured AI output (tags, text, time ranges, model provenance) so that other programs can index and search without a proprietary database. See the .vtag page for detail.

The product targets image, video, and audio with associated metadata paths your build supports. Exact codecs, containers, and file-size limits change between releases; start from the trial on your own samples for a definitive answer for your material.

VisionaryAI Suite is designed to complement NeoFinder workflows by producing structured AI metadata that can be indexed and reused, within the limits of your NeoFinder version and configuration. See NeoFinder workflow and your official product documentation for the step-by-step that applies to your build.

The architecture is aimed at serious collections, but throughput and limits always depend on your storage speed, GPU/CPU, model choices and batch settings. We avoid stating a number here that could be wrong the moment hardware changes. The trial and a pilot are the right next steps to measure your case.

The suite is built to swap and combine model components over time, so a static list in marketing would go stale. Your build’s about or credits window, and release documentation, are the right place to see the exact engines and versions in your installer.

You should assume some network use is possible (for example activation or updates) unless a specific build is documented as offline-capable. Core analysis of files on disk is designed to run without your media being uploaded to our servers as a default workflow—but confirm your contract and the exact product version’s privacy statement for a binding answer.

Both. Individuals and small teams can start with a trial; universities, broadcasters, agencies and public-sector archives are natural fits when scale, compliance, and multi-seat licensing enter the picture. Pricing is structured to grow with that reality.

Open the Download page for system requirements (Windows, NVIDIA recommended), storage expectations, and the ZIP link. If a host URL changes, contact us and we can send a current link.

Use the contact form or email info@bomarkanalys.se. For the fastest response while the web form is being connected, email is recommended.