nvidia/DeepSeek-V4-Flash-NVFP4

Model Overview

Description:

The NVIDIA DeepSeek-V4-Flash-NVFP4 model is a quantized version of DeepSeek AI's DeepSeek-V4-Flash model, an autoregressive Mixture-of-Experts language model that uses an optimized Transformer architecture with hybrid attention (Compressed Sparse Attention and Heavily Compressed Attention) and Manifold-Constrained Hyper-Connections. For more information, refer to the DeepSeek-V4-Flash model card. The NVIDIA DeepSeek-V4-Flash-NVFP4 model is quantized with Model Optimizer.

This model is ready for commercial/non-commercial use.

Third-Party Community Consideration

This model is not owned or developed by NVIDIA. This model has been developed and built to a third-party’s requirements for this application and use case; see link to Non-NVIDIA (DeepSeek-V4-Flash) Model Card.

References

Nvidia Model Optimizer: https://github.com/NVIDIA/Model-Optimizer

License/Terms of Use:

MIT

Deployment Geography:

Global

Use Case:

DeepSeek V4 is well-suited for advanced reasoning, agentic AI applications, tool use scenarios, and complex problem-solving in domains such as mathematics, software engineering, and enterprise AI assistants.

Release Date:

Hugging Face 05/28/2026 via https://huggingface.co/nvidia/DeepSeek-V4-Flash-NVFP4

Model Architecture:

Architecture Type: Transformers
Network Architecture: Mixture-of-Experts (MoE) with Hybrid Attention (Compressed Sparse Attention + Heavily Compressed Attention)
Number of Model Parameters: 284B in total and 13B activated

Input:

Input Type(s): Text
Input Format(s): String
Input Parameters: 1D (One-Dimensional): Sequences
Other Properties Related to Input: Supports multi-turn conversations with system prompts, user messages, and assistant responses. Maximum context length of 1 million tokens. Uses a custom encoding pipeline (encoding_dsv4) with three reasoning modes: Non-think (fast), Think High (logical analysis), and Think Max (full reasoning extent).

Output:

Output Type(s): Text
Output Format: String
Output Parameters: One-Dimensional (1D): Sequences
Other Properties Related to Output: Supports structured JSON output, function/tool calling, and reasoning content when enabled.

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:

Supported Runtime Engine(s):

• SGLang
  
• vLLM
  

Supported Hardware Microarchitecture Compatibility:

• NVIDIA Blackwell
  

Preferred Operating System(s):

• 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.

Model Version(s):

The model is DeepSeek-V4-Flash NVFP4 quantized with nvidia-modelopt v0.44.0

Training and Evaluation Datasets:

Calibration Dataset:

Link: cnn_dailymail, Nemotron-Post-Training-Dataset-v2
Data Collection Method by dataset: Automated.
Labeling Method by dataset: Automated.
Properties: The cnn_dailymail dataset is an English-language dataset containing just over 300k unique news articles as written by journalists at CNN and the Daily Mail. The Nemotron-Post-Training-Dataset-v2 is a post-training dataset curated by NVIDIA containing multi-turn conversations across diverse topics.

Training Datasets:

Data Modality: Undisclosed
Data Collection Method by dataset: Undisclosed
Labeling Method by dataset: Undisclosed
Data Size : Undisclosed
Properties: Undisclosed

Evaluation Dataset:

Datasets: AA-LCR, τ²-Bench Telecom, SciCode, IFBench
Data Collection Method by dataset: Hybrid: Automated, Human
Labeling Method by dataset: Hybrid: Human, Automated
DataSet Properties: We evaluated the model on long-context recall, agentic tool-use, coding, and instruction-following benchmarks: AA-LCR (Artificial Analysis Long Context Recall) evaluates a model's ability to accurately retrieve and recall information from long input contexts; τ²-Bench Telecom evaluates agentic tool-use and policy-adherence capabilities in dual-control telecom customer-service scenarios where the model interacts with a simulated user and external tools to resolve account issues; SciCode evaluates scientific coding capabilities; IFBench is a benchmark for evaluating instruction-following capabilities across diverse and structured task constraints.

Inference:

Acceleration Engine: SGLang, vLLM
Test Hardware: NVIDIA Blackwell B200

Post Training Quantization

This model was obtained by quantizing the weights and activations of DeepSeek-V4-Flash to NVFP4 data type, ready for inference with SGLang and vLLM. Only the weights and activations of the linear operators within transformer blocks in MoE are quantized.

Usage

Deploy with vLLM

To deploy the quantized NVFP4 checkpoint with vLLM, use the following command (you need 8xB200 GPUs):

python -m vllm.entrypoints.cli.main serve \
    nvidia/DeepSeek-V4-Flash-NVFP4 \
    --tensor-parallel-size 8 \
    --trust-remote-code \
    --kv-cache-dtype fp8 \
    --served-model-name nvfp4

Deploy with SGLang

Requires SGLang PR #25820. The integration auto-detects NVFP4 from the checkpoint's hf_quant_config.json (weights are stored in FP8 with "moe_quant_algo": "NVFP4"):

python3 -m sglang.launch_server --model nvidia/DeepSeek-V4-Flash-NVFP4 --tensor-parallel-size 8 --trust-remote-code

Evaluation

The accuracy benchmark results are presented in the table below:

Precision	AA-LCR	τ²-Bench Telecom	SciCode	IFBench
Baseline (ours)	0.658	0.943	0.481	0.788
NVFP4	0.655	0.942	0.481	0.795

Benchmarked with temperature=1.0, top_p=1.0, max num tokens 384000

Model Limitations: The base model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.

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 internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.

Users are responsible for model inputs and outputs. Users are responsible for ensuring safe integration of this model, including implementing guardrails as well as other safety mechanisms, prior to deployment.

Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns here.