Latest / Chips, compute and energy
NVIDIA releases Nemotron 3.5 retrieval models for agent systems
See how this story connectsNVIDIA released five Nemotron 3.5 models for embedding, reranking and document extraction: Nemotron 3.5 Embed 8B BF16, Embed 1B BF16, Embed 1B NVFP4, Rerank 8B BF16, and Parse at 631M parameters. The embedding models retrieve evidence from text and document images, including tables, charts and infographics. Text retrieval was evaluated across 34 languages. Nemotron 3.5 Embed 8B BF16 scored 61.17 nDCG@10 on ViDoRe v3.1, ranking first among dense retrieval models as of 20 October 2026.
Why it matters
In NVIDIA’s own agent tests, adding Nemotron 3.5 Rerank cut agent token cost per query by about 14 per cent with Embed 8B and 13 per cent with Embed 1B, while improving retrieval accuracy.
- When
- 21 Oct 2026
- ID
WTA-N-2026-0351- Where
- International
- Who
- NVIDIA
Summary by wheresthe.ai, from Hugging Face. Checked against the source on 21 Oct 2026.
Source copy: download the page as saved on 21 Oct 2026 (PDF). Also held by the Internet Archive.
Cite this story
WTA-N-2026-0351: “NVIDIA releases Nemotron 3.5 retrieval models for agent systems.” wheresthe.ai, 21 Oct 2026. https://wheresthe.ai/d/intl-nvidia-releases-nemotron-3-5-retrieval-models-for-agent-2026-10/ (source: Hugging Face).
Spotted an error? Email [email protected].