⏩

Ready to get started? Go to our Quickstart guide.

What we do

Loci helps creative teams make their 3D catalogs searchable by automatically tagging their assets. These tags—which include captions, metadata, and custom nouns—allow assets to seamlessly show up in search results.

We do this by creating world-class computer vision models, making them available through seamless APIs, and supporting teams to deploy them securely and privately.

---

Tagging with Loci

• Generative Tagging (Auto Tag):Our core feature, Generative Tagging, rapidly generates descriptive tags for 3D assets at scale. Generative Tagging also extracts detailed metadata such as polycount, hexadecimal values for colors, texture details, and animation specifications.
• Classification (Auto Categorize):Our Auto Categorize endpoint allows teams to use their own taxonomies to tag assets. We also offer a selection of pre-configured taxonomies for classifying objects, materials and styles. This can be useful for DAMS and marketplaces that already have a set of filters in place, or for creative teams with assets that require custom descriptors.
• Metadata (Auto Extract):Our Auto Extract endpoint extracts key metadata from any asset. It makes sorting assets easy by counting their polygons, vertices, nodes, and more.
• Color Classification:Our Color Classification endpoint determines the color composition of a 3D asset.
• Caption (Auto Caption):Our Auto Caption instantly generates a descriptive caption for any asset. It describes complex assets with rich, semantic text — perfect for search and gen-AI datasets.
• Similarity Search: This allows you to search through an index of your assets to instantly find geometric and visual matches to a 3D asset input.

---

Our mission

We help teams make the most of their creative hours.

We know that 3D is the future, but that 3D assets are hard to work with. Most search engines don't recognize 3D files, so finding the right one in a catalog of thousands can be painful. Creators spend countless hours sifting through DAMs or recreating lost work.

Tags are what make 3D assets searchable. Search engines may not recognize 3D files, but they can read text. When we assign 3D assets simple descriptive text tags, we allow search engines to look past the assets and search the tags instead.

So a search query for an "old tree" will return a Blender file of a withering oak—because the asset was tagged, and those tags showed up in search.

Many teams know this, and resolve to tagging all the assets in their catalog by hand. But this costs a lot of time and money. Data from one of our AAA clients reveals that manually tagging 100k assets can cost 6,000 human hours and $200,000 USD in labor. In other words, losing assets is painful, but so is tagging them.

At Loci, we've developed computer vision models that tag 3D assets automatically, precisely, and at scale. So that you can make your catalog searchable in minutes. And never need to hand-tag an asset again.

We're a team of 3D creatives and machine learning scientists who believe in solving this specific problem exceptionally well. Because 3D is the future, and great work deserves to be found.

Getting started

⏩

Ready to use our Public API? Go to our Quickstart guide.

1. Choose your product. Decide between using Loci's Generative Tagging, setting up Custom Classification, or any of our other available endpoints. Not sure which is the right fit for you? Visit our Documentation for technical details or hop on a short chat with us by Contacting Sales
2. Start tagging. Begin generating rapid, accurate tags for your 3D assets.
3. Support: Receive white-glove support as you integrate and tag.