LLM EXPLORER 58,297 MODELS INDEXED

Gemma 3n E4B by google

By google · 1975 downloads
HF Repo ↗ Copy URL

Gemma 3n E4B is an open-source language model by google. Features: 7.8b LLM, VRAM: 15.8GB, License: gemma, LLM Explorer Score: 0.22.

  Arxiv:1705.03551   Arxiv:1903.00161   Arxiv:1904.09728   Arxiv:1905.07830   Arxiv:1905.10044   Arxiv:1907.10641   Arxiv:1911.01547   Arxiv:1911.11641   Arxiv:2009.03300   Arxiv:2107.03374   Arxiv:2108.07732   Arxiv:2210.03057   Arxiv:2311.12022   Arxiv:2403.07974   Arxiv:2411.19799   Arxiv:2502.12404   Arxiv:2502.21228   Audio-text-to-text   Automatic-speech-recognition   Automatic-speech-translation   Endpoints compatible   Gemma3n   Image-text-to-text   Region:us   Safetensors   Sharded   Tensorflow   Video-text-to-text
Model Card on HF 🤗: https://huggingface.co/google/gemma-3n-E4B 

Gemma 3n E4B Parameters and Internals

LLM NameGemma 3n E4B
Repository 🤗https://huggingface.co/google/gemma-3n-E4B 
Model Size7.8b
Required VRAM15.8 GB
Updated2026-08-13
Maintainergoogle
Model Typegemma3n
Model Files  3.1 GB: 1-of-4   5.0 GB: 2-of-4   5.0 GB: 3-of-4   2.7 GB: 4-of-4
Model ArchitectureGemma3nForConditionalGeneration
Licensegemma
Transformers Version4.53.0.dev0
Tokenizer ClassGemmaTokenizer
Padding Token<pad>
Torch Data Typebfloat16

Quantized Models of the Gemma 3n E4B

Model
Likes
Downloads
VRAM
Gemma 3n E2B 4bit4534 GB
Gemma 3n E4B It 4bit72335 GB
Gemma 3n E4B It 8bit2379 GB
Gemma 3n E2B 8bit2216 GB

Best Alternatives to Gemma 3n E4B

Best Alternatives
Context / RAM
Downloads
Likes
Gemma 3n E4B It0K / 15.8 GB35121919
MedraN E4B0K / 15.8 GB50
...mma 3n E4B It Unsloth Bnb 4bit0K / 9.8 GB16499
Gemma 3n E4B Unsloth Bnb 4bit0K / 9.8 GB1894
Note: green Score (e.g. "73.2") means that the model is better than google/gemma-3n-E4B.