Qwen3.5-9B-GGUF

Qwen3.5-9B-GGUF

Unsloth's GGUF quantized copy of Qwen3.5-9B, a mid-size model that reads images and text and answers in text, for people running models locally.

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: Qwen3.5-9B-GGUF is a GGUF copy of Qwen3.5-9B, made by Unsloth. The base model reads images and text and replies in text. The record lists 9B parameters, and the 4-bit weights come to about 4.5 GB by its estimate.

The Qwen3.5 family is a causal language model with a vision encoder. It lists support for 201 languages and dialects, and a context length of 262,144 tokens natively.

Who it suits: It suits people who want a capable multimodal model on their own computer, without sending files to a hosted service. A file of about 4.5 GB needs a machine with some free memory.

Industry | Who uses it | How they use it | Example
Software | Developers | Run a local model for text and image questions | Asking what a screenshot shows
Education | Students and teachers | Practice and explain material with an offline model | Explaining a diagram from a textbook
Office work | Analysts | Summarize documents and read charts locally | Describing a chart image in words
Research | Hobbyist researchers | Try a recent model on their own hardware | Comparing answers across languages

Skip it if you want a hosted chat service and do not run models yourself.

Where to use it:

• Local apps and libraries that load GGUF files
• Windows, macOS or Linux, with Python for the download step
• Your own computer, with no hosted service needed

Use case | Who | What they get back
Ask about an image | Anyone with a local setup | A written answer about the picture
Long document questions | Researchers | Answers from a context window of 262,144 tokens
Multilingual chat | Global teams | Replies in many of the 201 languages the card lists

How to set it up: You need a few things in place first.

• Python and pip, for the download command
• A local app that can load GGUF files
• About 4.5 GB of free disk space for the 4-bit weights

1. Install the Hugging Face hub package with pip.
2. Download the repository with the hf command.
3. Open the GGUF file in your local app. The model card has the run details.

pip install -U huggingface_hub
hf download unsloth/Qwen3.5-9B-GGUF

How to use it: Once the file is loaded in your local app, you chat with it or send it images with a question. Training your own variant on your machine is also possible through Unsloth, the maker of this copy.

Example: You load the file and attach a photo of a chart with the question "What does this chart show?" The model answers in text. It describes what the picture contains and what the question needs.

For newcomers and experts: Start small. Download the file, open it in a GGUF-capable app and ask one plain question before you try images. If it runs well, add a picture.

Qwen3.5 mixes Gated DeltaNet layers with gated attention layers, in a pattern repeated 8 times across 32 layers. The card calls this a hybrid design that keeps inference fast. The GGUF copy trades some precision for a much smaller file. Unsloth says its Dynamic 2.0 quant method keeps accuracy high, and the card gives a native context of 262,144 tokens that can stretch to 1,010,000.

Download the GGUF file with hf
Load it in a local app
Send text or an image
The model replies in text

Good to know:

• License: Apache 2.0
• Cost to run: none named, but you need a machine with room for about 4.5 GB of weights
• Data it reaches: it downloads files from Hugging Face onto your computer

On Hugging Face, 13 November 2026

• Task: image text to text
• Library: transformers
• Parameters: 9B
• Weights: about 4.5 GB at 4-bit (estimate)
• Formats: GGUF
• Downloads in the last 30 days: 1,969,901
• Likes: 1,074
• License: apache-2.0
• Based on: Qwen/Qwen3.5-9B
• Quantized versions: 5
• Fine-tunes: 2
• Tags: transformers, gguf, unsloth, image-text-to-text, conversational

Links

• Model: https://huggingface.co/unsloth/Qwen3.5-9B-GGUF
• Files: https://huggingface.co/unsloth/Qwen3.5-9B-GGUF/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download unsloth/Qwen3.5-9B-GGUF

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