USD Search lets developers, creators, and workflow specialists search large collections of OpenUSD and 3D assets using natural-language descriptions or reference images — without any manual tagging. It covers both flat-list search across an asset library and structural navigation within individual scenes:
- Search by description — natural-language queries surface matching USD assets, renders, and reference images even when filenames don't contain the keywords.
- Search by example — supply a reference image to find visually similar 3D assets. Useful for "more like this" or matching a target look.
- Search inside scenes — spatial queries within a USD file, prim and property filters, and forward / reverse dependency lookups across the asset graph.
- Auto-generated metadata — captions, tags, materials, and style are produced for every asset during ingestion. The metadata schema is fully configurable.
- Quality-checked results — irrelevant matches are filtered out so what surfaces actually matches the query.
- Connect to your storage — point USD Search at S3, a Storage API, or Omniverse Nucleus; assets are discovered and indexed at scale.
Try it yourself against the public NVIDIA-hosted instance at https://search.simready.omniverse.nvidia.com with a single shell command — no install:
./scripts/quickstart.sh --hosted --query "yellow forklift"Tip
If you're using Claude Code or Codex in this repo, you can ask in plain language instead — "find a yellow forklift" — and the agent dispatches the right skill. See docs/agent-skills.md for the full set.
Deploy the USD Search stack on your own hardware with one command — the search / info / asset-graph APIs plus Swagger at /docs/:
docker compose -f docker-compose.yml -f docker-compose.gpu-plugins.yml up -d --buildOpen http://localhost:8080 — by default / redirects to /docs/ (Swagger). After the stack reports healthy (~60s), ./scripts/quickstart-smoke.sh exercises every gateway-proxied endpoint.
| Component | Required version |
|---|---|
| OS | Linux |
| Docker Compose | v2.26 or newer |
| NVIDIA GPU | with nvidia-container-toolkit configured |
To run without a GPU (SigLIP2 on CPU, no renderer), drop the -f docker-compose.gpu-plugins.yml flag.
For the full local-deployment guide — VLM auto-tagging, custom S3 buckets, local-filesystem assets, Nucleus, Explorer WebUI — see docs/local-deployment.md.
Tip
Guided setup with an agent: Claude Code or Codex can walk through the same flow interactively (storage backend, GPU/VLM plugins, credentials), then hand back to a sample query. See docs/agent-skills.md.
Scalable deployment (Kubernetes): for production, a Kubernetes cluster is required. USD Search ships a Helm chart at helm/usdsearch/, also published to the NGC Catalog. See helm/usdsearch/README.md for the full installation guide.
Licensed under the Apache License, Version 2.0. Third-party component licenses are listed in THIRD_PARTY_NOTICE.md.
This project is currently not accepting contributions. See CONTRIBUTING.md.
Please report security vulnerabilities per the policy in SECURITY.md.
This project adheres to the Contributor Covenant Code of Conduct. See CODE_OF_CONDUCT.md.