gemma-4-E2B-it-GGUF

gemma-4-E2B-it-GGUF

Unsloth's GGUF files of Google DeepMind's Gemma 4 E2B instruction-tuned model, which takes text and images and writes text, with a 128K context. For people who run models locally on laptops and small machines.

Made by unsloth on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: This repository holds Unsloth's GGUF files of gemma-4-E2B-it, an instruction-tuned model from Google DeepMind's Gemma 4 family. Gemma 4 models are multimodal, taking text and images as input and writing text. Audio is supported on the small models. The Hugging Face record lists 4.6B parameters and about 2.3 GB at 4-bit.

Who it suits: People who want a small multimodal model they can run on their own laptop or edge device, and developers who need reasoning, coding or function calling in a compact model.

Industry | Who uses it | How they use it | Example
Software | Developers on small machines | Run a multimodal model locally | Ask questions about a screenshot
Agents | Builders of tool-using agents | Use native function calling | Let the model call a weather or search function
On-device apps | Teams shipping to laptops and phones | Keep inference on the device | Answer a user's question with no server call

You can skip it if you need the biggest Gemma 4 sizes, since the family also has 26B A4B and 31B versions.

Where to use it:

• Unsloth Studio, which the card says can run and fine-tune Gemma 4
• llama.cpp and other GGUF-based tools
• Windows, macOS or Linux with Python, to download the files

Use case | Who | What they get back
Image questions | App developers | Text answers about an image
Coding help | Developers | Code and explanations
Tool-calling agents | Agent builders | Function calls from the model's native support
Structured chat | Prompt writers | Replies shaped by a system prompt

How to set it up: You need a GGUF-capable tool and enough memory for a roughly 2.3 GB file at 4-bit. The card points to Unsloth's guide for running Gemma 4, and it notes a re-download for the latest chat template.

• Python and pip
• A GGUF-capable runtime such as llama.cpp or Unsloth Studio
• Disk space for the file you choose

1. Download the GGUF files.
2. Open the guide the card links for run steps.
3. Load a file in your chosen tool.

pip install -U huggingface_hub
hf download unsloth/gemma-4-E2B-it-GGUF

How to use it: Load the model in your tool, then chat with text or images. Gemma 4 supports the system role, so you can set the behavior in a system prompt. Thinking modes can be configured.

Example: You attach a photo of a receipt and ask, "What is the total?". The model replies with the amount as text. The card does not show a sample output in the part used here.

For newcomers and experts: If you are new, try it in Unsloth Studio, where the card says the model runs with tool calling. Start with a short text question before you add images.

For experts, Gemma 4 spans dense and mixture-of-experts designs in four sizes: E2B, E4B, 26B A4B and 31B. The small models have a 128K context window and the medium ones 256K. Unsloth later added MTP support and updated the files for Google's latest chat template.

Text or image input
Gemma 4 E2B, quantized to GGUF
Text output

Good to know:

• Family sizes: E2B, E4B, 26B A4B and 31B
• Context: 128K tokens for the small models
• Updates: re-download for the latest chat template and llama.cpp fixes
• License: Apache 2.0
• Cost to run: no API key or paid service named
• Data it reaches: downloads files from Hugging Face, nothing else is named
• Not for: the card names no limits, and the output is text only

On Hugging Face, 13 November 2026

• Task: image text to text
• Parameters: 4.6B
• Weights: about 2.3 GB at 4-bit (estimate)
• Formats: GGUF
• Downloads in the last 30 days: 674,565
• Likes: 356
• License: apache-2.0
• Based on: google/gemma-4-E2B-it
• Quantized versions: 5
• Fine-tunes: 2
• Tags: gguf, gemma4, unsloth, gemma, google, image-text-to-text, conversational

Links

• Model: https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF
• Files: https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download unsloth/gemma-4-E2B-it-GGUF

Then follow the usage in its model card. Nothing in this kit runs by itself.