NVIDIA releases open embedding models that take first place on the RTEB leaderboard
The company has launched Nemotron 3 Embed, a set of open and commercially available models intended to boost retrieval quality while offering practical deployment paths for production RAG, agentic retrieval, code search, and agent memory.
The release contains three open weights. An 8B model leads the pack by topping the RTEB leaderboard. Two 1B variants follow, built for production environments where cost and latency are priorities.
Model options and targets
- Nemotron-3-Embed-8B-BF16: The flagship quality anchor. This model ranks #1 on RTEB and suits precision-critical retrieval and high-stakes enterprise RAG.
- Nemotron-3-Embed-1B-BF16: The high-efficiency standard. Designed for production retrieval where latency and cost matter.
- Nemotron-3-Embed-1B-NVFP4: A hardware-accelerated variant. Optimised for Blackwell to deliver high-throughput retrieval with a smaller memory footprint.
Features and deployment
Beyond the leaderboard result, the release includes a production-ready feature set for enterprise retrieval deployments.
- Open weights, datasets, and recipes: Teams can inspect, tune, fine-tune, and deploy retrieval models on their own infrastructure.
- 32k context window: Supports retrieval over long documents, large code contexts, and multi-turn agent histories while reducing truncation.
- Multilingual and code retrieval: Handles global enterprise data, technical documentation, and multi-file code repositories.
- NVIDIA NVFP4 efficiency: Provides a Blackwell-optimised 4-bit deployment path for high-throughput retrieval with a smaller memory footprint.
- Fine-tuning and distillation recipes: NVIDIA NeMo AutoModel recipes support domain adaptation and model compression for teams adapting retrieval models to their own data.
- Day-0 ecosystem integration: Available immediately on Hugging Face, deployable as an NVIDIA NIM microservice, supported by vLLM, and accessible through leading AI cloud and inference partners.
Evaluation results
The team evaluated Nemotron 3 Embed across retrieval quality, downstream agentic efficiency, and deployment tradeoffs. The 8B model establishes the quality ceiling, while the 1B BF16 and NVFP4 variants bring the same retrieval-focused design to lower-cost and higher-throughput settings.
RTEB leadership and benchmark gains
The first evaluation occurred on RTEB, where Nemotron-3-Embed-8B-BF16 ranks #1. Testing also covered ViDoRe V3 Text, MMTEB Retrieval, and LongEmbed using average NDCG@10.
- Nemotron-3-Embed-8B-BF16 ranks #1 on RTEB, scoring 78.5% on RTEB and 75.5% on MMTEB Retrieval.
- Nemotron-3-Embed-1B-BF16 brings much of the 8B model’s retrieval quality into a smaller deployment footprint. It scores 72.4% on RTEB, reducing error rate by 27% over its 1B predecessor (llama-nemotron-embed-vl-1b-v2), and scores 71.0% on MMTEB Retrieval, reducing error rate by 28%.
Impact on agentic performance
To evaluate retrieval in an agentic setting, the team used a search agent powered by Nemotron 3 Ultra and varied the embedding model used by the retrieval system. Better retrieval can return relevant evidence earlier, helping the agent avoid repeated searches, unnecessary reasoning turns, and extra context inspection. The comparison measured average retrieval accuracy against estimated downstream agentic token cost per query across ViDoRe V3, BRIGHT, and BrowseComp-Plus.
The search agent uses Nemotron 3 Ultra. Downstream token cost is estimated from Nemotron 3 Ultra input/output token counts using the GPT-5.5 pricing formula.
Stronger retrieval reduces downstream agentic token cost. More accurate retrievers return relevant evidence earlier, which helps agents complete tasks with fewer repeated searches and fewer reasoning turns. In these evaluations, the Nemotron 3 Embed models improve the agentic retrieval frontier, with the 8B model delivering both the highest average retrieval accuracy and the lowest estimated downstream token cost across ViDoRe V3, BRIGHT, and BrowseComp-Plus.
Scaling retrieval with NVFP4 on Blackwell
For high-throughput deployments, teams often choose smaller embedding models to meet latency and cost targets. Nemotron-3-Embed-1B-NVFP4 is designed to narrow the gap between serving efficiency and retrieval quality by using native NVFP4 acceleration on NVIDIA Blackwell architectures. The model quantises the weights and activations of linear layers to NVFP4 for efficient inference, and uses Quantization-Aware Distillation (QAD) to help recover accuracy for long input sequences.
- Serving Efficiency: NVFP4 on Blackwell delivers up to 2x higher throughput than BF16 for high-throughput, low-latency retrieval serving.
- Accuracy retention: The NVFP4 variant retains 99%+ of BF16 retrieval accuracy while reducing memory footprint.
Production NIM performance
For production-scale retrieval systems, the serving stack also needs to preserve that efficiency under real request loads, across different input sequence lengths and hardware targets. To make Nemotron 3 Embed performant at enterprise scale today, the team is releasing an optimised NVIDIA NIM microservice for the 1B model. As shown in the data, the Rust-based Nemotron 3 Embed NIM matches or outperforms the vLLM checkpoint on NVIDIA GB200 and RTX PRO 6000 GPUs across ISLs of 256 and 1024.
How the models were built
Nemotron-3-Embed-8B-BF16 adapts the Ministral-3-8B-Instruct-2512 backbone by converting its causal decoder into a bidirectional encoder for full-sequence retrieval. The model is trained with contrastive pre-training on a blend of web-sourced and synthetic text pairs, then fine-tuned on curated multilingual retrieval datasets across domains such as legal, finance, medical, business, and education. This 8B model serves as the flagship embedding model, while earlier 8B teacher checkpoints from the same development line were used to distil the efficient 1B variants.
Scaling down to 1B
The 1B model is not a small retriever trained from scratch. The team first applied the bidirectional adaptation recipe to the Ministral-3-3B-Instruct-2512 backbone to establish a 3B retriever base, then compressed it through two rounds of structured pruning and distillation.
First, the 3B parent model was compressed to a 2B intermediate footprint using NVIDIA ModelOpt’s mcore_minitron Neural Architecture Search engine. The NAS pipeline searched across hidden width, FFN size, attention heads, and depth under a strict parameter budget to identify an efficient architecture for retrieval workloads.
The resulting 2B intermediate model was then distilled from an 8B teacher checkpoint to recover ranking accuracy. The team used a combined cosine distance loss and mean squared error loss on a multilingual, in-domain retrieval data blend to align the student’s embeddings with the teacher.
This same sequence, ModelOpt structured pruning followed by 8B teacher distillation, was repeated a second time to compress the 2B intermediate model down to the final 1.14B embedding model. Final training used a progressive two-stage context-scaling schedule:
- Stage 1: Focused on broad multilingual alignment at 1024-token context length to reconstruct the core retrieval behavior of the parent model.
- Stage 2: Expanded context length to 4096 tokens and added long-context synthetic and reasoning datasets, helping the 1B model retain discriminative recall across longer inputs.
Technical specifications
The following table summarises the core technical specifications and deployment targets for the Nemotron 3 Embed models.
Model specs
- Nemotron-3-Embed-8B-BF16: 8.0B parameters, 4096 embedding dimension, 32k context window, Mean pooling, query: / document: input prefix, General GPU Inference.
- Nemotron-3-Embed-1B-BF16: 1.14B parameters, 2048 embedding dimension, 32k context window, Mean pooling, query: / document: input prefix, Low-latency CPU/GPU.
- Nemotron-3-Embed-1B-NVFP4: 1.14B parameters, 2048 embedding dimension, 32k context window, Mean pooling, query: / document: input prefix, NVIDIA Blackwell/GB200.
Enterprise partner evaluations
Enterprise ISVs, AI-native companies, and memory providers are already evaluating Nemotron 3 Embed across agentic retrieval, agent memory, code retrieval, and production inference workflows.
“Context is the key to agentic accuracy. Our Context Intelligence Graph uses embeddings and semantic similarity to deliver the most relevant enterprise context to agents like EnterpriseClaw, which we launched with NVIDIA in May. Early results from NVIDIA’s new Nemotron 3 Embed models are promising, particularly for question answering, where they show improvements over our current model. We’re excited about their potential to further improve the accuracy and reliability of our enterprise agents.” – Adi Kuruganti, Chief AI and Development Officer at Automation Anywhere
“Our initial evaluation of the NVIDIA Nemotron 3 Embed models shows strong retrieval performance for our agentic retrieval use cases. The availability of both 1B and 8B variants gives teams the flexibility to balance quality, latency, and deployment requirements across different environments. We’re excited to continue evaluating the models and exploring how they can support high-performance retrieval for production AI applications.” – Mani Gill, Senior Vice President of Product Management at Boomi
IBM has seen promising early results evaluating the new NVIDIA Nemotron Embed model in a proof-of-concept built on watsonx.
