What are the hardware requirements for GLM 5.2 NVFP4?

GLM 5.2 NVFP4 requires approximately 429.9 GB of VRAM and supports a context window of 1024K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed GLM 5.2 NVFP4 and how large is it?

GLM 5.2 NVFP4 is developed by nvidia, a model with 381b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate GLM 5.2 NVFP4?

GLM 5.2 NVFP4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

GLM 5.2 NVFP4 by nvidia — VRAM 429.9GB, 1024K context

Name: GLM 5.2 NVFP4
Author: nvidia

GLM 5.2 NVFP4 is an open-source language model by nvidia. Features: 381b LLM, VRAM: 429.9GB, Context: 1024K, License: mit, MoE, LLM Explorer Score: 0.44.

4-bit precision 8-bit Base model:quantized:zai-org/g... Base model:zai-org/glm-5.2 Conversational Fp4 Glm-5 Glm moe dsa Model optimizer Modelopt Moe Nvidia Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/nvidia/GLM-5.2-NVFP4

GLM 5.2 NVFP4 Parameters and Internals

LLM Name	GLM 5.2 NVFP4
Repository 🤗	https://huggingface.co/nvidia/GLM-5.2-NVFP4
Base Model(s)	zai-org/GLM-5.2 zai-org/GLM-5.2
Model Size	381b
Required VRAM	429.9 GB
Updated	2026-07-23
Maintainer	nvidia
Model Type	glm_moe_dsa
Model Files	10.0 GB: 1-of-47 10.0 GB: 2-of-47 10.0 GB: 3-of-47 10.0 GB: 4-of-47 10.0 GB: 5-of-47 10.0 GB: 6-of-47 10.0 GB: 7-of-47 10.0 GB: 8-of-47 10.0 GB: 9-of-47 10.0 GB: 10-of-47 10.0 GB: 11-of-47 10.0 GB: 12-of-47 10.0 GB: 13-of-47 10.0 GB: 14-of-47 10.0 GB: 15-of-47 10.0 GB: 16-of-47 10.0 GB: 17-of-47 10.0 GB: 18-of-47 10.0 GB: 19-of-47 10.0 GB: 20-of-47 10.0 GB: 21-of-47 10.0 GB: 22-of-47 10.0 GB: 23-of-47 10.0 GB: 24-of-47 10.0 GB: 25-of-47 10.0 GB: 26-of-47 10.0 GB: 27-of-47 10.0 GB: 28-of-47 10.0 GB: 29-of-47 10.0 GB: 30-of-47 10.0 GB: 31-of-47 10.0 GB: 32-of-47 10.0 GB: 33-of-47 10.0 GB: 34-of-47 10.0 GB: 35-of-47 10.0 GB: 36-of-47 10.0 GB: 37-of-47 10.0 GB: 38-of-47 10.0 GB: 39-of-47 10.0 GB: 40-of-47 10.0 GB: 41-of-47 9.9 GB: 42-of-47 10.0 GB: 43-of-47
Model Architecture	GlmMoeDsaForCausalLM
License	mit
Context Length	1048576
Model Max Length	1048576
Transformers Version	5.11.0
Vocabulary Size	154880

Rank the GLM 5.2 NVFP4 capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to GLM 5.2 NVFP4

Best Alternatives	Context / RAM	Downloads	Likes
SuperGLM 5.2 Abliterated NVFP4	1024K / 410 GB	2520	49
GLM 5.2 NVFP4	1024K / 429.9 GB	167	3