BREAKING
NVIDIA Launches Nemotron 3 Embed
0
8B on LMEB
0
1B on LMEB
RTEB and MMTEB Scores
8B RTEB
78.5
1B RTEB
72.4
8B MMTEB
75.5
1B MMTEB
71
Two Sizes, One Trade-Off
8B Model
Accuracy
●
4096-dim embeddings
●
32k context, 34 languages
●
Top of RTEB
1B Model
Efficiency
●
2048-dim vectors
●
Blackwell NVFP4 build
●
Up to 2x throughput
0
%
Mem0 Retrieval@10
0
languages
Open Weights on Hugging Face
AI NEWS BLITZ
NVIDIA has released Nemotron 3 Embed, a new family of open-weight text embedding models.