Flow - 4x faster than typing
Mini toolsNemotron 3 Embed 1B NVFP4
By NVIDIA
Model family: Nemotron
Nemotron-3-Embed-1B-NVFP4 is the NVFP4 quantized version of Nemotron-3-Embed-1B-BF16, a 1.14B parameter multilingual text embedding model built on a Ministral-3-3B-Instruct based pruned encoder with hidden size 2048. Quantized using NVIDIA Model Optimizer to 8-bit precision (FP8_E4M3), it is optimized for ultra high throughput retrieval serving on Blackwell class hardware while sharing the same embedding space as the BF16 model, allowing largely interchangeable use after validating retrieval quality. It processes text inputs up to 32768 tokens and outputs dense vector embeddings for multilingual retrieval, semantic similarity, and RAG. It is ready for commercial use and positioned as the hardware accelerated deployment variant in the Nemotron 3 Embed collection, targeting massive scale infrastructure.
New
Text
Overview
Nemotron-3-Embed-1B-NVFP4 is a hardware accelerated, 8-bit quantized variant of NVIDIA's Nemotron-3-Embed-1B-BF16 text embedding model. It shares the same embedding space and produces dense multilingual text embeddings for retrieval and RAG, optimized for high throughput, massive scale inference on Blackwell GPUs.
About NVIDIA
Industry:
Computer Hardware Manufacturing
Company Size:
42000
Location:
Santa Clara, California, US
Website:
nvidia.com
Last updated: July 17, 2026
