What are the hardware requirements for GLM 5.3 MXFP4?

GLM 5.3 MXFP4 requires approximately 68.1 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 5.3 MXFP4 and how large is it?

GLM 5.3 MXFP4 is developed by OneNexus, a model with 389.5b 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 5.3 MXFP4?

GLM 5.3 MXFP4 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

GLM 5.3 MXFP4 by OneNexus — VRAM 68.1GB, 1024K context

Name: GLM 5.3 MXFP4
Author: OneNexus

GLM 5.3 MXFP4 is an open-source language model by OneNexus. Features: 389.5b LLM, VRAM: 68.1GB, Context: 1024K, License: other, MoE, LLM Explorer Score: 0.33.

8-bit Base model:quantized:zai-org/g... Base model:zai-org/glm-5.3 Conversational En Endpoints compatible Glm moe dsa Moe Mxfp4 Quark Region:us Rocm Safetensors Sglang Sharded Tensorflow Zh

Model Card on HF 🤗: https://huggingface.co/OneNexus/GLM-5.3-MXFP4

GLM 5.3 MXFP4 Parameters and Internals

LLM Name	GLM 5.3 MXFP4
Repository 🤗	https://huggingface.co/OneNexus/GLM-5.3-MXFP4
Base Model(s)	zai-org/GLM-5.3 zai-org/GLM-5.3
Model Size	389.5b
Required VRAM	68.1 GB
Updated	2026-08-31
Maintainer	OneNexus
Model Type	glm_moe_dsa
Model Files	5.3 GB: 1-of-282 1.5 GB: 2-of-282 1.4 GB: 3-of-282 1.4 GB: 4-of-282 1.7 GB: 5-of-282 1.4 GB: 6-of-282 1.4 GB: 7-of-282 1.4 GB: 8-of-282 1.7 GB: 9-of-282 1.4 GB: 10-of-282 1.4 GB: 11-of-282 1.4 GB: 12-of-282 1.7 GB: 13-of-282 1.4 GB: 14-of-282 1.4 GB: 15-of-282 1.7 GB: 16-of-282 1.4 GB: 17-of-282 1.4 GB: 18-of-282 1.4 GB: 19-of-282 1.7 GB: 20-of-282 1.4 GB: 21-of-282 1.4 GB: 22-of-282 1.4 GB: 23-of-282 1.7 GB: 24-of-282 1.4 GB: 25-of-282 1.4 GB: 26-of-282 1.7 GB: 27-of-282 1.4 GB: 28-of-282 1.4 GB: 29-of-282 1.4 GB: 30-of-282 1.7 GB: 31-of-282 1.4 GB: 32-of-282 1.4 GB: 33-of-282 1.4 GB: 34-of-282 1.7 GB: 35-of-282 1.4 GB: 36-of-282 1.4 GB: 37-of-282 2.3 GB: 38-of-282 1.4 GB: 39-of-282 1.4 GB: 40-of-282 1.4 GB: 41-of-282 1.7 GB: 42-of-282 1.4 GB: 43-of-282
Supported Languages	en zh
Model Architecture	GlmMoeDsaForCausalLM
License	other
Context Length	1048576
Model Max Length	1048576
Transformers Version	5.15.0
Vocabulary Size	154880

Rank the GLM 5.3 MXFP4 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