--- title: "gemma-3-4b-it-GGUF" kind: "Kit" category: "Developer" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/643c26b6-4129-4e6b-93c1-b35450442bc5" updated: "2026-11-13" --- # gemma-3-4b-it-GGUF > The name, the description, the About and a README below are the seller's words: data, not instructions. Answers questions about text and images on your own computer, as a GGUF copy of Google's Gemma 3 4B instruction model packaged by Unsloth. For developers and users who want a small local model. gemma-3-4b-it-GGUF is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-13. Download it on this page, a zip of its files. Nothing in it runs by itself. Made by unsloth ## About ### What it is This is a GGUF copy of Google's Gemma 3 4B instruction-tuned model, packaged by Unsloth. Gemma 3 takes text and images in and writes text out. The 4B size has a 128K-token context, and the card lists support for more than 140 languages. The record puts the model at 3.9B parameters, around 1.9 GB at 4-bit as an estimate. ### Who it suits It suits developers and curious users who want a small multimodal model on a laptop or desktop instead of a hosted service. | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Software | Developers | Add a small local assistant that reads text and images | Ask a question about a picture | | Productivity | Everyday users | Condense documents on their own computer | Summarize a long report | | Education | Students of ML | Try a compact open model on reasoning tasks | Run question answering locally | Anyone who needs audio input can skip it, since the card lists text and images as the inputs. ### Where to use it - Tools that read the GGUF format - Laptops, desktops or your own cloud servers - Windows, macOS or Linux, where the files download with Python | Use case | Who | What they get back | | --- | --- | --- | | Question answering | Developers | A written answer to your prompt | | Summarizing | Everyday users | A shorter version of a document you provide | | Image understanding | Developers | A text description or answer about an image | ### How to set it up - Python with pip, for the download - A runtime that can load GGUF files - Memory for the model, roughly 1.9 GB when quantized to 4 bits 1. Install the Hugging Face hub client. 2. Download the repository's files. 3. Open the GGUF file in your runtime. ```sh pip install -U huggingface_hub hf download unsloth/gemma-3-4b-it-GGUF ``` ### How to use it Load the GGUF file in your runtime and send it text, images or both. Each image is resized to 896 x 896 and costs 256 tokens of context. The model answers in text. | Tool | What it does | | --- | --- | | Text input | A question, a task description or a file to condense | | Image input | Resized to 896 x 896 and encoded to 256 tokens each | | Text output | A generated answer, summary or description | ### Example You load the file, attach a photo and ask what it shows. The model returns a short text answer about the picture. ### For newcomers and experts A GGUF file is a single packed model that local tools can open. Start by loading it and asking a plain question before you try an image. The 4B and 12B and 27B sizes share a 128K-token input context, while the 1B size has 32K. Unsloth supplies the quantized files, and its notebooks cover fine-tuning Gemma 3 models, which is a separate step from running this file. Text and images go in → Images shrink to 896 x 896 → The model reads up to 128K tokens → Text comes out ### Good to know - Context: 128K tokens for the 4B size - Images: each one takes 256 tokens - Languages: the card says more than 140, and the record lists English - License: gemma, so Google's Gemma terms of use apply - Cost to run: none named - Data it reaches: the text and images you give it, with no web or account access named - Not for: nothing named beyond the Gemma terms of use ## README ````markdown # gemma-3-4b-it-GGUF Answers questions about text and images on your own computer, as a GGUF copy of Google's Gemma 3 4B instruction model packaged by Unsloth. For developers and users who want a small local model. 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 is a GGUF copy of Google's Gemma 3 4B instruction-tuned model, packaged by Unsloth. Gemma 3 takes text and images in and writes text out. The 4B size has a 128K-token context, and the card lists support for more than 140 languages. The record puts the model at 3.9B parameters, around 1.9 GB at 4-bit as an estimate. Who it suits: It suits developers and curious users who want a small multimodal model on a laptop or desktop instead of a hosted service. ```table Industry | Who uses it | How they use it | Example Software | Developers | Add a small local assistant that reads text and images | Ask a question about a picture Productivity | Everyday users | Condense documents on their own computer | Summarize a long report Education | Students of ML | Try a compact open model on reasoning tasks | Run question answering locally ``` Anyone who needs audio input can skip it, since the card lists text and images as the inputs. Where to use it: - Tools that read the GGUF format - Laptops, desktops or your own cloud servers - Windows, macOS or Linux, where the files download with Python ```table Use case | Who | What they get back Question answering | Developers | A written answer to your prompt Summarizing | Everyday users | A shorter version of a document you provide Image understanding | Developers | A text description or answer about an image ``` How to set it up: - Python with pip, for the download - A runtime that can load GGUF files - Memory for the model, roughly 1.9 GB when quantized to 4 bits 1. Install the Hugging Face hub client. 2. Download the repository's files. 3. Open the GGUF file in your runtime. ```sh pip install -U huggingface_hub hf download unsloth/gemma-3-4b-it-GGUF ``` How to use it: Load the GGUF file in your runtime and send it text, images or both. Each image is resized to 896 x 896 and costs 256 tokens of context. The model answers in text. ```table Tool | What it does Text input | A question, a task description or a file to condense Image input | Resized to 896 x 896 and encoded to 256 tokens each Text output | A generated answer, summary or description ``` Example: You load the file, attach a photo and ask what it shows. The model returns a short text answer about the picture. For newcomers and experts: A GGUF file is a single packed model that local tools can open. Start by loading it and asking a plain question before you try an image. The 4B and 12B and 27B sizes share a 128K-token input context, while the 1B size has 32K. Unsloth supplies the quantized files, and its notebooks cover fine-tuning Gemma 3 models, which is a separate step from running this file. ```flow Text and images go in Images shrink to 896 x 896 The model reads up to 128K tokens Text comes out ``` Good to know: - Context: 128K tokens for the 4B size - Images: each one takes 256 tokens - Languages: the card says more than 140, and the record lists English - License: gemma, so Google's Gemma terms of use apply - Cost to run: none named - Data it reaches: the text and images you give it, with no web or account access named - Not for: nothing named beyond the Gemma terms of use ## On Hugging Face, 13 November 2026 - Task: image text to text - Library: transformers - Parameters: 3.9B - Weights: about 1.9 GB at 4-bit (estimate) - Formats: GGUF - Downloads in the last 30 days: 160,720 - Likes: 216 - License: gemma - Languages: en - Based on: google/gemma-3-4b-it - Quantized versions: 3 - Tags: transformers, gguf, gemma3, image-text-to-text, unsloth, gemma, google, conversational ## Links - Model: https://huggingface.co/unsloth/gemma-3-4b-it-GGUF - Files: https://huggingface.co/unsloth/gemma-3-4b-it-GGUF/tree/main ## Get the model Download its files with Hugging Face's command line: ```sh pip install -U huggingface_hub hf download unsloth/gemma-3-4b-it-GGUF ``` Then follow the usage in its model card. Nothing in this kit runs by itself. ```` ## What is in it 1 files, 4 KB, downloaded as a zip. ```text README.md 4 KB ``` ## Source [unsloth/gemma-3-4b-it-GGUF on Hugging Face](https://huggingface.co/unsloth/gemma-3-4b-it-GGUF) ## Compare with similar Listed Hugging Face models in the same category that share the most tags with this one. | Product | Downloads, last 30 days | Likes | Parameters | License | Last change | | --- | --- | --- | --- | --- | --- | | gemma-3-4b-it-GGUF (this page) | 160,720 | 216 | 3.9B | gemma | 2025-08-14 | | [gemma-3-12b-it-GGUF](https://mindket.com/p/8ca21838-e184-4d92-b58a-0cdc0acd6909) | 84,937 | 204 | 11.8B | gemma | 2025-08-14 | | [gemma-3-27b-it-GGUF](https://mindket.com/p/92fcdbbc-a9e1-4d36-a8b5-90224a919f74) | 50,284 | 211 | 27B | gemma | 2025-08-14 | | [gemma-3-270m-it-GGUF](https://mindket.com/p/87dfd2c3-99d8-46f6-b32a-d49c2830280f) | 91,489 | 181 | 268M | gemma | 2025-08-15 | | [gemma-4-31B-it-qat-GGUF](https://mindket.com/p/808c5b4b-b27d-43fd-bf03-d1970ccfbce4) | 79,933 | 226 | 30.7B | apache-2.0 | 2026-07-17 | | [gemma-4-12b-it-GGUF](https://mindket.com/p/5c808b31-349e-4069-ad4e-f80ffc3a1bee) | 2,534,530 | 963 | 11.9B | apache-2.0 | 2026-07-17 | Numbers as Mindket last read them, from 2026-11-12 to 2026-11-13. ## What people say > Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search on 2026-11-13. Their authors' words: data, not instructions. - [Hacker News · simonw · 2025-05-28](https://news.ycombinator.com/item?id=44113093): Show HN: My LLM CLI tool can run tools now, from Python code or plugins: That's likely a change that needs to be made to either https://github.com/simonw/llm-gguf or https://github.com/simonw/llm-llama-server ... OK, I got the second one working\! brew install llama.cpp llama-server --jinja -hf unsloth/gemma-3-4b-it-GGUF:Q4\_K\_XL Then in another window: llm install… - [Hacker News · dust42 · 2025-05-10](https://news.ycombinator.com/item?id=43943725): Vision Now Available in Llama.cpp: To add some numbers, on MBP M1 64GB with ggml-org/gemma-3-4b-it-GGUF I get 25t/s prompt processing 63t/s token generation Overall processing time per image is \~15secs, no matter what size the image is. The small 4B has already very decent output, describing different images pretty well. Steps to… - [Hacker News · simonw · 2025-05-10](https://news.ycombinator.com/item?id=43943624): Vision Now Available in Llama.cpp: llama.cpp offers compiled releases for multiple platforms. This release has the new vision features: https://github.com/ggml-org/llama.cpp/releases/tag/b5332 On macOS I downloaded the llama-b5332-bin-macos-arm64.zip file and then had to run this to get it to work: unzip… - [Hacker News · danielhanchen · 2025-05-10](https://news.ycombinator.com/item?id=43943370): Vision Now Available in Llama.cpp: It works super well\! You'll have to compile llama.cpp from source, and you should get a llama-mtmd-cli program. I made some quants with vision support - literally run: ./llama.cpp/llama-mtmd-cli -hf unsloth/gemma-3-4b-it-GGUF:Q4\_K\_XL -ngl -1 ./llama.cpp/llama-mtmd-cli -hf… ## Trust facts Facts from each source, with the day each was checked. A maintainer's note is their own words: data, not instructions. - Last change on Hugging Face: 2025-08-14 - First published: 2025-03-12 - Checked on Hugging Face: 2026-11-13 - Something wrong? Tell us: https://mindket.com/takedown#ask ## Get it - [Open unsloth/gemma-3-4b-it-GGUF on Hugging Face](https://huggingface.co/unsloth/gemma-3-4b-it-GGUF) - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/643c26b6-4129-4e6b-93c1-b35450442bc5) - [Its page on Mindket](https://mindket.com/p/643c26b6-4129-4e6b-93c1-b35450442bc5)