New Models
NVIDIA releases Nemotron 3.5 Embed, Rerank and Parse, open retrieval models that search charts and scanned pages for AI agents
NVIDIA added five open, commercially usable models to its Nemotron Retriever family on October 21, 2026: multimodal embedding models at 8B and 1B, an 8B reranker and a 631M-parameter document parser. NVIDIA says they top…