LLM EXPLORER 62,007 MODELS INDEXED

GLM 5.2 Int4 Int8Mix by QuantTrio

By QuantTrio · 71011 downloads
HF Repo ↗ Copy URL

GLM 5.2 Int4 Int8Mix is an open-source language model by QuantTrio. Features: 785b LLM, VRAM: 123.5GB, Context: 1024K, License: mit, MoE, LLM Explorer Score: 0.33.

  Arxiv:2602.15763   Arxiv:2603.12201 Base model:quantized:zai-org/g...   Base model:zai-org/glm-5.2   Compressed-tensors   Conversational   Endpoints compatible   Glm moe dsa   Int4   Int8   Moe   Region:us   Safetensors   Sharded   Tensorflow   Vllm   W4a16   W8a16
Model Card on HF 🤗: https://huggingface.co/QuantTrio/GLM-5.2-Int4-Int8Mix 

GLM 5.2 Int4 Int8Mix Parameters and Internals

LLM NameGLM 5.2 Int4 Int8Mix
Repository 🤗https://huggingface.co/QuantTrio/GLM-5.2-Int4-Int8Mix 
Base Model(s)  zai-org/GLM-5.2   zai-org/GLM-5.2
Model Size785b
Required VRAM123.5 GB
Updated2026-07-23
MaintainerQuantTrio
Model Typeglm_moe_dsa
Model Files  1.9 GB: 1-of-124   3.2 GB: 2-of-124   3.2 GB: 3-of-124   3.2 GB: 4-of-124   3.2 GB: 5-of-124   3.2 GB: 6-of-124   3.2 GB: 7-of-124   3.2 GB: 8-of-124   3.2 GB: 9-of-124   3.2 GB: 10-of-124   3.2 GB: 11-of-124   3.2 GB: 12-of-124   3.2 GB: 13-of-124   3.2 GB: 14-of-124   3.2 GB: 15-of-124   3.2 GB: 16-of-124   3.2 GB: 17-of-124   3.2 GB: 18-of-124   3.2 GB: 19-of-124   3.2 GB: 20-of-124   3.2 GB: 21-of-124   3.2 GB: 22-of-124   3.2 GB: 23-of-124   3.2 GB: 24-of-124   3.2 GB: 25-of-124   3.2 GB: 26-of-124   3.2 GB: 27-of-124   3.2 GB: 28-of-124   3.2 GB: 29-of-124   3.2 GB: 30-of-124   3.2 GB: 31-of-124   3.2 GB: 32-of-124   3.2 GB: 33-of-124   3.2 GB: 34-of-124   3.2 GB: 35-of-124   3.2 GB: 36-of-124   3.2 GB: 37-of-124   3.2 GB: 38-of-124   3.2 GB: 39-of-124
Model ArchitectureGlmMoeDsaForCausalLM
Licensemit
Context Length1048576
Model Max Length1048576
Transformers Version5.12.0
Vocabulary Size154880