LLM EXPLORER 57,918 MODELS INDEXED

EGM 4B by nvidia

By nvidia · 194 downloads
HF Repo ↗ Copy URL

EGM 4B is an open-source language model by nvidia. Features: 4b LLM, VRAM: 27.2GB, License: apache-2.0, LLM Explorer Score: 0.24.

Base model:finetune:qwen/qwen3... Base model:qwen/qwen3-vl-4b-th...   Conversational   Deploy:azure   En   Endpoints compatible   Grpo   Image-text-to-text   Multimodal   Qwen3-vl   Qwen3 vl   Region:us   Safetensors   Sharded   Tensorflow   Visual-grounding
Model Card on HF 🤗: https://huggingface.co/nvidia/EGM-4B 

EGM 4B Parameters and Internals

LLM NameEGM 4B
Repository 🤗https://huggingface.co/nvidia/EGM-4B 
Base Model(s)  Qwen3 VL 4B Thinking   Qwen/Qwen3-VL-4B-Thinking
Model Size4b
Required VRAM27.2 GB
Updated2026-08-10
Maintainernvidia
Model Typeqwen3_vl
Model Files  5.0 GB: 1-of-2   5.0 GB: 1-of-4   4.7 GB: 2-of-2   5.0 GB: 2-of-4   4.9 GB: 3-of-4   2.6 GB: 4-of-4
Supported Languagesen
Model ArchitectureQwen3VLForConditionalGeneration
Licenseapache-2.0
Model Max Length8192
Transformers Version4.57.1
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace

Best Alternatives to EGM 4B

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3 VL 4B Instruct0K / 8.9 GB3611570437
Qwen3 VL 4B Instruct FP80K / 6.1 GB30110365
Qwen3 VL 4B Thinking0K / 8.9 GB21715118
OMEGA 4B SpatialThink 08040K / 8.9 GB161
EGM 4B SFT0K / 8.9 GB11391
Qwen3 VL 4B Instruct0K / 8.9 GB1786411
CapRL Qwen3VL 4B0K / 9.7 GB118212
GELab Zero 4B Preview0K / 8.9 GB235154
ZwZ 4B0K / 9.7 GB15532
Qwen3 VL 4B Thinking FP80K / 6.1 GB272330
Note: green Score (e.g. "73.2") means that the model is better than nvidia/EGM-4B.