8B multimodal reranking model that scores how relevant a text or document-image passage is to a query.
Nemotron-3.5-Rerank-8B-BF16 was developed by NVIDIA for multilingual and multimodal question-answering retrieval. It is a world-class reranker optimized to produce a logit representing the relevance of a document to a text query. Documents may contain text, page images, or both, including charts, tables, and infographics. The model can be integrated into classical or agentic vision RAG pipelines. For text retrieval, the model was evaluated across 34 languages: English, Arabic, Assamese, Bengali, Bulgarian, Chinese, Danish, Dutch, Finnish, French, German, Hindi, Hinglish, Indonesian, Italian, Japanese, Korean, Malay, Marathi, Nepali, Norwegian, Persian, Portuguese, Romanian, Russian, Spanish, Swahili, Swedish, Tamil, Telugu, Thai, Ukrainian, Urdu, and Vietnamese. For image retrieval, it was evaluated across 18 languages: Arabic, Bengali, Chinese, English, Persian, Finnish, French, German, Hindi, Indonesian, Japanese, Korean, Russian, Spanish, Swahili, Telugu, Thai, and Yoruba. The model may support additional languages beyond these evaluation sets.
This model is ready for commercial or non-commercial use.
GOVERNING HOSTING TERMS: Use of this trial service is governed by the NVIDIA API Trial Terms of Service. Use of the model is governed by OpenMDW License Agreement, version 1.1. Additional Information: Apache License, Version 2.0 for Ministral-3-8B-Instruct-2512-BF16.
Global
Use Case: The reranking model serves as a key component of a multimodal retrieval system, such as a vision RAG pipeline, where it helps improve overall retrieval accuracy. A multimodal retrieval system typically uses a multimodal embedding model (dense retrieval) to return a set of potentially relevant documents given an input query. The reranking model then evaluates these candidate documents in greater detail and reorders them based on their relevance to the query, producing a final ranking. By refining the initial retrieval results, the reranking stage can improve the precision of the retrieved documents and ultimately lead to more accurate downstream responses.
Build.NVIDIA.com: 10/20/2026 (planned) via model page
Hugging Face: 10/20/2026 via model card
References:
Architecture Type: Transformer
Network Architecture: The released model contains a Pixtral-based vision encoder and a bidirectional Ministral3 language-model tower. Training proceeded in two stages. In the first stage, an 8B reranker checkpoint derived from the Ministral-3-8B-Instruct-2512-BF16 backbone was trained on a blend of text-only datasets. In the second stage, the 8B reranker model initialized from the first-stage checkpoint was trained on image and combined image-text datasets.
Total Parameters: 8.38B
Input Types: Text, Image
Input Formats: Text: String; Image: Red, Green, Blue (RGB)
Input Parameters: Text: One-Dimensional (1D), Image: Two-Dimensional (2D)
Other Input Properties: Each input is one text query paired with one document. The document may be an image, text, or a combination of image and text. The default image longest_edge is 1,120 pixels and can be increased to 1,540. The maximum context length is 32,768 tokens.
Output Types: Floats
Output Formats: List of Floats
Output Parameters: One-Dimensional (1D)
Other Output Properties: Each output value corresponds to a raw logit. The higher the value, the more relevant a document is to a given query.
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.
Runtime Engines:
Supported Hardware Microarchitecture Compatibility:
Preferred Operating Systems: Linux
The integration of foundation and fine-tuned models into AI systems requires additional testing using use-case-specific data to ensure safe and effective deployment. Following the V-model methodology, iterative testing and validation at both unit and system levels are essential to mitigate risks, meet technical and functional requirements, and ensure compliance with safety and ethical standards before deployment.
Nemotron-3.5-Rerank-8B-BF16
Data Modality: Text, Image
Image Training Data Size: 2.3M samples containing images, with or without accompanying text
Text Training Data Size: 21M text-only data samples
Training Data Collection: Automated, Manually-Labeled, Synthetic
Training Labeling: Hybrid: Automated, Manually-Labeled, Synthetic
Training Properties: Training was conducted using publicly available, commercially permissible datasets and synthetically generated datasets.
The language model component of the Nemotron-3.5-Rerank-8B-BF16 was trained on a blend of text-only dataset files. The public source files used to create text-only training datasets are listed in the Nemotron-3-Embed-8B-BF16 model card.
The following publicly available datasets were also used to train the vision encoder and multimodal projector on image-only and combined image-text inputs.
Synthetic text and image datasets were created either by generating queries from seed documents or by generating complete question–answer pairs through LLM-based prompting using the models listed below.
| LLMs used to generate synthetic datasets |
|---|
| nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16 nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4 |
| Qwen/Qwen3-Next-80B-A3B-Instruct Qwen/Qwen3-235B-A22B Qwen/Qwen3.5-397B-A17B Qwen/Qwen3.5-35B-A3B Qwen/Qwen3.6-27B Qwen/Qwen3.6-35B-A3B Qwen/Qwen3.8-27B |
| google/gemma-4-31B-it |
| mistralai/Pixtral-12B-2409 |
| openai/gpt-oss-120b openai/gpt-oss-20b |
| 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 |
| Digital Corpora | https://digitalcorpora.org/corpora/file-corpora/cc-main-2021-31-pdf-untruncated/ |
| Vdr-multilingual | https://huggingface.co/datasets/llamaindex/vdr-multilingual-train |
Testing Data Collection: Not Applicable
Testing Labeling: Not Applicable
Testing Properties: Not Applicable. Model quality was assessed using the evaluation benchmark datasets described in the Evaluation Dataset subsection.
Benchmark:
Benchmark Score: NDCG@10
The Nemotron-3.5-Rerank-8B-BF16 model was evaluated with the model's default image longest_edge set to 1,120 pixels. Qwen3-VL-Reranker-8B was evaluated with the model's default max_pixels set to 1,310,720. Both models reranked the top 50 candidates retrieved by the Nemotron-3.5-Embed-8B-BF16 model.
| Model | ViDoRe V3 | MIRACL-VISION | ||
|---|---|---|---|---|
| image | text | image + text | image | |
| Nemotron-3.5-Embed-8B-BF16 | 60.39 | 60.65 | 61.47 | 70.15 |
| Qwen3-VL-Reranker-8B | 66.32 | 65.30 | 66.49 | 72.75 |
| Nemotron-3.5-Rerank-8B-BF16 | 67.49 | 67.35 | 68.09 | 76.35 |
Evaluation Data Collection: Automated, Manually-Labeled, Synthetic
Evaluation Labeling: Hybrid: Automated, Manually-Labeled, Synthetic
Evaluation Properties: ViDoRe V3 is a visual document retrieval benchmark covering 8 public and 2 private datasets across professional domains. MIRACL-VISION is a multilingual visual retrieval dataset for 18 different languages. It is an extension of MIRACL, a popular text-only multilingual retrieval dataset.
Acceleration Engine: vLLM
Test Hardware: NVIDIA Lovelace (L40S)
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. Developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
Please make sure you have proper rights and permissions for all input image and video content; if image or video includes people, personal health information, or intellectual property, the image or video generated will not blur or maintain proportions of image subjects included.
For more detailed information on ethical considerations for this model, please see the Model Card++ Bias, Explainability, Privacy, and Safety & Security Subcards.
Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns here.