--- title: "nemotron-3-embed-1b" publisher: "nvidia" type: "endpoint" updated: "2026-07-16T19:07:19.287Z" description: "1B embedding model for semantic search, retrieval, and RAG applications." canonical: "https://build.nvidia.com/nvidia/nemotron-3-embed-1b" --- # Model Overview ### Description: **Nemotron-3-Embed-1B-BF16** is a versatile text embedding model trained by NVIDIA and optimized for retrieval and semantic similarity tasks. It provides strong multilingual and cross-lingual retrieval capabilities and is designed to serve as a foundational component in text-based Retrieval-Augmented Generation (RAG) systems. This model was evaluated on 34 languages: English, Arabic, Assamese, Bengali, Bulgarian, Chinese, Danish, Dutch, Finnish, French, German, Hindi, Hinglish, Indonesian, Italian, Japanese, Korean, Malay, Marathi, Nepalese, Norwegian, Persian, Portuguese, Romanian, Russian, Spanish, Swahili, Swedish, Tamil, Telugu, Thai, Ukrainian, Urdu, Vietnamese. The model generates dense vector embeddings from multilingual text inputs, enabling retrieval, semantic search, and (agentic) RAG workflows. As a core component of text retrieval systems, an embedding model transforms text, such as questions or passages, into dense vector representations. These models are typically transformer encoders that process input tokens and produce embeddings suitable for efficient similarity matching. Among models of comparable size, **Nemotron-3-Embed-1B-BF16** achieves state-of-the-art performance across multiple multilingual retrieval benchmarks. *This model is ready for commercial use.* ### License/Terms of Use: **GOVERNING TERMS:** The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). Use of this model is governed by the [OpenMDW License Agreement, version 1.1 (OpenMDW-1.1)](https://openmdw.ai/license/1-1/). Additional Information: Built with [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) which is released under Apache 2.0. **You are responsible for ensuring that your use of NVIDIA provided models complies with all applicable laws.** **Model Developer:** NVIDIA ### Deployment Geography: **Global** ### Use Case: The **Nemotron-3-Embed-1B-BF16** is most suitable for users who want to build a multilingual question-and-answer application over a large text corpus, leveraging the latest dense retrieval technologies, including RAG pipelines. ### Release Date: 07/16/2026 via [Build.NVIDIA.com](https://build.nvidia.com/nvidia/nemotron-3-embed-1b/) ## Model Architecture: **Architecture Type:** Transformer
**Network Architecture:** [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) based pruned model
**Embedding Dimension:** 2048
**Max Sequence Length:** 32768
**Number of Model Parameters:** ~1.14B
**Precision:** bf16 The **Nemotron-3-Embed-1B-BF16** was derived from the **Nemotron-3-Embed-3B** text-embedding model through two iterative rounds of structured pruning and distillation, using NVIDIA ModelOpt mcore_minitron Neural Architecture Search (NAS). ## Input(s): **Input Type(s):** Text
**Input Format(s):** String / List of strings
**Input Parameters:** One Dimensional (1D)
**Other Properties Related to Input:** Text inputs longer than the maximum context length of 32768 tokens should be truncated or chunked. ## Output(s): **Output Type(s):** Floats (dense vector embeddings)
**Output Format(s):** List of floats
**Output Parameters:** One Dimensional (1D)
**Other Properties Related to Output:** The model outputs a 2048-dimensional embedding vector for each input text string. __Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA's hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions.__ ## Software Integration: **Runtime Engines:** Rust + CUDA **Supported Hardware Microarchitecture Compatibility:**
NVIDIA Ampere
NVIDIA Blackwell
NVIDIA Hopper
NVIDIA Lovelace
**Preferred/Supported Operating System(s):** Linux
**Supported GPU SKUs:** NVIDIA H100 80GB HBM3, NVIDIA A100 SXM4 80GB, NVIDIA L40S, NVIDIA A10G, NVIDIA GB200, NVIDIA RTX PRO 6000 Blackwell Server Edition ## Performance: | Model Name | RTEB 16 | ViDoRE-V3 text | MMTEB (Retrieval) | | --- | --- | --- | --- | | llama-nemotron-embed-1b-v2 | 60.47 | 52.10 | 59.58 | | llama-nemotron-embed-vl-1b-v2 | 61.98 | 52.54 | 59.71 | | **Nemotron-3-Embed-1B-BF16** | **72.38** | **57.76** | **71.05** | *Avg. NDCG@10 on text retrieval benchmarks (chunk retrieval), evaluated at sequence length 4096.* ## Model Version(s): Nemotron-3-Embed-1B-BF16 Short Name: `nemotron-3-embed-1b` ## Training, Testing, and Evaluation Datasets: ### Dataset Overview: **Total Size:** 8.5M+ data samples
**Total Number of Datasets:** 161 dataset files
**Dataset Partition:** Training [100%], Testing [N/A — evaluation benchmarks used separately], Validation [N/A — evaluation benchmarks used separately]. Model distillation training was conducted using publicly available, commercially permissible datasets and synthetically generated datasets. Synthetic data was created either by generating queries from seed documents or by generating complete question–answer pairs through LLM-based prompting. ### Public Datasets: | Dataset name | Reference | | --- | --- | | MIRACL | [https://huggingface.co/datasets/miracl/miracl](https://huggingface.co/datasets/miracl/miracl) | | MLDR | [https://huggingface.co/datasets/Shitao/MLDR](https://huggingface.co/datasets/Shitao/MLDR) | | HotpotQA | [https://hotpotqa.github.io/](https://hotpotqa.github.io/) | | NQ | [https://huggingface.co/datasets/sentence-transformers/embedding-training-data](https://huggingface.co/datasets/sentence-transformers/embedding-training-data) | | Squad | [https://rajpurkar.github.io/SQuAD-explorer/](https://rajpurkar.github.io/SQuAD-explorer/) | | Stack Exchange | [https://archive.org/details/stackexchange](https://archive.org/details/stackexchange) | | Hover | [https://hover-nlp.github.io/](https://hover-nlp.github.io/) | | TAT-QA | [https://huggingface.co/datasets/next-tat/TAT-QA](https://huggingface.co/datasets/next-tat/TAT-QA) | | FinQA | [https://github.com/czyssrs/FinQA/tree/main](https://github.com/czyssrs/FinQA/tree/main) | | PubMedQA | [https://huggingface.co/datasets/qiaojin/PubMedQA/viewer/pqa_labeled](https://huggingface.co/datasets/qiaojin/PubMedQA/viewer/pqa_labeled) | | MedQuAD | [https://github.com/abachaa/MedQuAD](https://github.com/abachaa/MedQuAD) | | JaQuAD | [https://huggingface.co/datasets/SkelterLabsInc/JaQuAD](https://huggingface.co/datasets/SkelterLabsInc/JaQuAD) | | coir_apps | [https://huggingface.co/datasets/CoIR-Retrieval/apps](https://huggingface.co/datasets/CoIR-Retrieval/apps) | | coir_cosqa | [https://huggingface.co/datasets/CoIR-Retrieval/cosqa](https://huggingface.co/datasets/CoIR-Retrieval/cosqa) | | coir_stackoverflow_qa | [https://huggingface.co/datasets/CoIR-Retrieval/stackoverflow-qa](https://huggingface.co/datasets/CoIR-Retrieval/stackoverflow-qa) | | coir_codetrans_dl | [https://huggingface.co/datasets/CoIR-Retrieval/codetrans-dl](https://huggingface.co/datasets/CoIR-Retrieval/codetrans-dl) | | coir_codetrans_contest | [https://huggingface.co/datasets/CoIR-Retrieval/codetrans-contest](https://huggingface.co/datasets/CoIR-Retrieval/codetrans-contest) | | synthetic_text2sql | [https://huggingface.co/datasets/CoIR-Retrieval/synthetic-text2sql](https://huggingface.co/datasets/CoIR-Retrieval/synthetic-text2sql) | | SWE-bench | [https://huggingface.co/datasets/princeton-nlp/SWE-bench/viewer/default/train](https://huggingface.co/datasets/princeton-nlp/SWE-bench/viewer/default/train) | | MLQA | [https://github.com/facebookresearch/MLQA](https://github.com/facebookresearch/MLQA) | | SpartQA | [https://github.com/HLR/SpartQA_generation](https://github.com/HLR/SpartQA_generation) | | Winogrande | [https://github.com/allenai/winogrande](https://github.com/allenai/winogrande) | | TempReason | [https://huggingface.co/datasets/tonytan48/TempReason](https://huggingface.co/datasets/tonytan48/TempReason) | ### Synthetic Datasets: Synthetic query-document pairs were generated either from scratch or by using seed datasets to generate queries with the models listed below.
LLMs used to generate synthetic datasets
Qwen/Qwen3-Next-80B-A3B-Instruct
Qwen/Qwen3-235B-A22B
Qwen/Qwen3.5-397B-A17B
Qwen/Qwen3.6-27B
Qwen/Qwen3.6-35B-A3B
google/gemma-4-31B-it
openai/gpt-oss-120b
openai/gpt-oss-20b
nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4
Seed Datasets
Dataset Reference
FinePdfs https://huggingface.co/datasets/HuggingFaceFW/finepdfs
CentralActs https://zenodo.org/records/5088102
BRIGHT https://huggingface.co/datasets/xlangai/BRIGHT
MultiHiertt https://github.com/psunlpgroup/MultiHiertt
### Training Dataset: **Data Modality:** Text
**Training Data Size:** 8.5M+
**Data Collection Method by dataset:** Hybrid: Human, Automated, Synthetic
**Labeling Method by dataset:** Hybrid: Human, Automated, Synthetic
**Properties:** Model training was conducted on text datasets using question–passage pairs from publicly available, commercially permissible datasets and synthetically generated datasets. ### Testing Dataset: **Properties:** Not Applicable. Model quality was assessed using the evaluation benchmark datasets described in the Evaluation Dataset subsection.
**Data Collection Method by dataset:** Not Applicable
**Labeling Method by dataset:** Not Applicable ### Evaluation Dataset: **Data Collection Method by dataset:** Hybrid: Human, Automated, Synthetic
**Labeling Method by dataset:** Hybrid: Human, Automated, Synthetic
**Properties:** This model is evaluated on 16 public tasks on [Retrieval Embedding Benchmark (RTEB)](https://huggingface.co/blog/rteb), a benchmark designed to reliably evaluate the retrieval accuracy of embedding models for real-world applications. More details on RTEB can be found on their [leaderboard](https://huggingface.co/spaces/mteb/leaderboard?benchmark_name=RTEB%28beta%29). The model was also evaluated on the [MMTEB (Retrieval) benchmark datasets](https://huggingface.co/spaces/mteb/leaderboard) ([paper](https://arxiv.org/pdf/2502.13595)), and on the eight text datasets (extracted via OCR) from [ViDoRe-V3 benchmark](https://mteb-leaderboard.hf.space/benchmark/ViDoRe(v3)). ## Inference: **Acceleration Engine:** Rust + CUDA
**Test Hardware:** NVIDIA Lovelace (L40S) ## Ethical Considerations: NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their supporting model team to ensure this model meets requirements for the relevant industry and use case, and address unforeseen product misuse. For more detailed information on ethical considerations for this model, please see the Model Card++ Bias, Explainability, Safety & Security, and Privacy Subcards. Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en-us/support/submit-security-vulnerability/). ## Capabilities - **Function Calling:** Not supported ## Bias | Field | Response | | ----- | ----- | | Participation considerations from adversely impacted groups [protected classes](https://www.senate.ca.gov/content/protected-classes) in model design and testing | None | | Measures taken to mitigate against unwanted bias | None | | Bias Metric (If Measured): | None | ## Explainability | Field | Response | | ----- | ----- | | Intended Task/Domain: | Passage and query embedding for question and answer retrieval | | Model Type: | Transformer encoder | | Intended Users: | Generative AI creators working with conversational AI models - users who want to build a multilingual question and answer application over a large text corpus, leveraging the latest dense retrieval technologies. | | Output: | Array of float numbers (Dense Vector Representation for the input text) | | Describe how the model works: | Model transforms the tokenized input text into a dense vector representation. | | Name the adversely impacted groups this has been tested to deliver comparable outcomes regardless of: | Not Applicable | | Technical Limitations & Mitigation: | The model's max sequence length is 32768. Therefore, the longer text inputs should be truncated. | | Verified to have met prescribed NVIDIA quality standards: | Yes | | Performance Metrics: | Accuracy, Throughput, and Latency | | Potential Known Risks: | This model does not always guarantee to retrieve the correct passage(s) for a given query. | | Licensing: | This model and its associated configuration files are licensed under the [OpenMDW License Agreement, version 1.1 (OpenMDW-1.1)](https://openmdw.ai/license/1-1/). Additional Information: Built with [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) which is released under Apache 2.0. | ## Privacy | Field | Response | | ----- | ----- | | Generatable or reverse engineerable personal data? | None | | Was consent obtained for any personal data used? | Not Applicable | | Personal data used to create this model? | None Known | | How often is the dataset reviewed? | Before Every Release | | Is there provenance for all datasets used in training? | Yes | | Does data labeling (annotation, metadata) comply with privacy laws? | Yes | | Was data from user interactions with the AI model (e.g. user input and prompts) used to train the model? | Yes | | Is data compliant with data subject requests for data correction or removal, if such a request was made? | No, not possible with externally-sourced data. | | Applicable Privacy Policy | https://www.nvidia.com/en-us/about-nvidia/privacy-policy/ | ## Safety & Security | Field | Response | | ----- | ----- | | Model Application(s): | Text Embedding for Retrieval | | Describe the life critical impact (if present). | Not Applicable | | Use Case Restrictions: | This model and its associated configuration files are licensed under the [OpenMDW License Agreement, version 1.1 (OpenMDW-1.1)](https://openmdw.ai/license/1-1/). Additional Information: Built with [Ministral-3-3B-Instruct-2512](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512) which is released under Apache 2.0. | | Model and dataset restrictions: | The Principle of least privilege (PoLP) is applied limiting access for dataset generation and model development. Restrictions enforce dataset access during training, and dataset license constraints adhered to. |