ToolsRecipesCompareDGX SparkModelsPeopleAboutSponsor

← All recipes · DGX Spark hub

Private document Q&A on a DGX Spark

Draft, untested Assembled from NVIDIA playbooks and owner reports. RAGShelf has not tested it end to end on a DGX Spark.

  1. Update DGX OS and firmware (Ollama's benchmarks used firmware 580.95.05; source).
  2. Install Ollama and Open WebUI with NVIDIA's Open WebUI with Ollama playbook.
  3. Pull gpt-oss:120b as the chat model and an embedding model such as bge-m3.
  4. In Open WebUI, create a knowledge collection, upload your PDFs and Office files, set the embedding model to the local bge-m3, and attach the collection to a model.
  5. Keep everything offline: make sure no cloud API keys are configured.

Alternative: AnythingLLM + Qdrant from amasu/dgx-spark-rag.

Last updated 2026-10-24

The RAGShelf weekly

New local RAG tools, releases, and tested setups. One email a week, no spam.