Model Overview

Description:

The NVIDIA GLM-5.2 NVFP4 model is the quantized version of ZAI’s GLM-5.2 model, which is an auto-regressive language model that uses an optimized transformer architecture. GLM-5.2 is a Mixture-of-Experts (MoE) model for reasoning and coding that uses sparse attention (with an IndexShare indexer) to support a long context. For more information, please check here. The NVIDIA GLM-5.2 NVFP4 model is quantized with Model Optimizer.

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

License/Terms of Use:

GOVERNING TERMS: Use of the model is governed by the MIT License, same as the base model.

Deployment Geography:

Global

Use Case:

Developers looking to take off-the-shelf, pre-quantized models for deployment in AI Agent systems, chatbots, RAG systems, and other AI-powered applications.

Release Date:

Hugging Face 06/25/2026 via https://huggingface.co/nvidia/GLM-5.2-NVFP4

References

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

Model Architecture:

Architecture Type: Transformers Network Architecture: GLM-5.2 (GlmMoeDsaForCausalLM) Number of Model Parameters: 753B in total and 40B activated

Input:

Input Type(s): Text Input Format(s): String Input Parameters: One-Dimensional (1D) Other Properties Related to Input: Context length up to 1M

Output:

Output Type(s): Text Output Format: String Output Parameters: 1D (One-Dimensional): Sequences Other Properties Related to Output: None

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 version is NVFP4 1.0 version and is quantized with nvidia-modelopt v0.46.0

Training, Testing, and Evaluation Datasets:

We performed evaluation using the benchmarks noted under Evaluation Datasets. We did not perform training or testing for this Model Optimizer release. The methods noted under Training and Testing Datasets below represent the data collection and labeling methods used by the third-party to train and test the underlying model.

Training Dataset:

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

Testing Dataset:

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

Evaluation Dataset:

Datasets: GPQA Diamond, SciCode, IFBench, AA-LCR, τ²-Bench Telecom Data Collection Method by dataset: Hybrid: Automated, Human Labeling Method by dataset: Hybrid: Human, Automated Properties: We evaluated the model on text-based reasoning, coding, long-context recall, and agentic tool-use benchmarks: GPQA Diamond contains 448 graduate-level multiple-choice questions written by domain experts in biology, physics, and chemistry; SciCode evaluates scientific coding capabilities; IFBench is a benchmark for evaluating instruction-following capabilities across diverse and structured task constraints; 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.

Inference:

Acceleration Engine: SGLang, vLLM Test Hardware: NVIDIA B200 NVIDIA B300

Post Training Quantization

This model was obtained by quantizing the weights and activations of GLM-5.2 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 experts are quantized. The shared expert is not quantized.

Usage

SGLang

This checkpoint was served with the latest SGLang image (lmsysorg/sglang:latest). GLM-5.2's glm_moe_dsa architecture requires transformers>=5.3.0, which we installed in the container before launching the server. You can also use lmsysorg/sglang:dev-glm52-nvfp4

pip install -U "transformers>=5.3.0" && \
python3 -m sglang.launch_server \
    --model nvidia/GLM-5.2-NVFP4 \
    --tensor-parallel-size 8 \
    --quantization modelopt_fp4 \
    --tool-call-parser glm47 \
    --reasoning-parser glm45 \
    --trust-remote-code \
    --chunked-prefill-size 16384 \
    --mem-fraction-static 0.80

vLLM

To serve this checkpoint with vLLM, use the vllm/vllm-openai:v0.23.0 image and run:

vllm serve nvidia/GLM-5.2-NVFP4 \
    --tensor-parallel-size 8 \
    --enable-expert-parallel \
    --trust-remote-code \
    --reasoning-parser glm45 \
    --tool-call-parser glm47 \
    --enable-auto-tool-choice \
    --kv-cache-dtype fp8_e4m3 \
    --host 0.0.0.0 --port 8000

Evaluation

The accuracy benchmark results are presented in the table below. AA-LCR was measured with SGLang; all other benchmarks were measured with vLLM.

Baseline: GLM-5.2-FP8. Benchmarked with temperature=1.0, top_p=0.95. GPQA Diamond used max_new_tokens=100000; all other benchmarks used max_new_tokens=64000.

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. 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, Safety & Security, and Privacy Subcards.

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

SUBCARDS:

Explainability

Bias

Safety & Security

Privacy