Alpie Seed 9B GGUF

Alpie Seed 9B GGUF

Runs Alpie Seed 9B, an image-and-text model from 169Pi AI, on your own machine through GGUF files in three bit depths. Suits developers who want local vision-language chat.

Made by 169Pi 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: These are GGUF quantized copies of Alpie Seed 9B, an open-weight vision-language model from 169Pi AI. The files are meant for local use on desktops, workstations and some phones. They come in three bit depths, so you can trade memory against quality.

Who it suits: Developers and researchers who want an image-and-text model running offline on their own hardware.

Industry | Who uses it | How they use it | Example
Software | Developers | Run a vision-language model locally | Ask a question about the contents of an image
Research | Researchers | Compare quantization levels for size and quality | Run the 2-bit and 4-bit builds on the same prompts
Mobile | App tinkerers | Run the model offline on a phone | The card shows a phone demo in PocketPal

If you want the full-precision model, you can skip these quantized copies.

Where to use it:

• llama.cpp on a desktop or workstation
• Ollama, with ready-made tags
• PocketPal on a phone
• Kaggle Models, where the model is also listed

Use case | Who | What they get back
Image understanding | Developers | Text answers about a picture, using the vision tag
Reasoning and coding | Students and developers | Help with math problems, code and debugging
Offline chat | Anyone with a capable device | A local assistant with no network needed

How to set it up:

• Python with pip, or Ollama installed
• Memory for the build you choose, from about 3.8 GB to 6.6 GB for the Ollama tags
• A runtime that reads GGUF, such as llama.cpp

1. Install the Hugging Face hub package and download the GGUF files, or skip this and use Ollama.
2. Pick a quantization that fits your memory.
3. Run it with llama.cpp or Ollama.

pip install -U huggingface_hub
hf download 169Pi/Alpie_Seed_GGUF

How to use it: With Ollama, run the default tag, or add a tag after the model name to choose a smaller build or the vision build.

ollama run 169pi/alpie-seed

ollama run 169pi/alpie-seed:q2
ollama run 169pi/alpie-seed:vision

Tool | What it does
alpie-seed:latest | The default build, about 5.6 GB, text input
alpie-seed:q2 | The smallest build, about 3.8 GB, text input
alpie-seed:q3 | A mid-size build, about 4.6 GB, text input
alpie-seed:vision | About 6.6 GB, takes text and images

Example: You start the vision tag and send a photo together with a question about what is in it. The model answers in text.

For newcomers and experts: If you are new, begin with the default Ollama tag and ask a plain question. Move to the vision tag when you want to send images.

Q2 uses the least memory, Q3 sits in the middle and Q4 keeps the most fidelity, so Q4 suits desktops and workstations. The base model has a native 262K context window, and the Ollama builds list 256K. Real sizes depend on the GGUF variant, the context length and the runtime.

Choose a quantization level
Get the GGUF files or an Ollama tag
Run it in llama.cpp or Ollama
Send text, or images with the vision tag

Good to know:

• Only the vision tag takes images, and the other listed Ollama tags take text
• Q2 suits memory-constrained devices
• License: Apache 2.0
• Cost to run: none named, you supply the hardware
• Data it reaches: the prompts and images you send, and the files you download
• Not for: sending images to the text-only tags

Title: Ollama download size
latest: 5.6 GB
q2: 3.8 GB
q3: 4.6 GB
vision: 6.6 GB

On Hugging Face, 13 November 2026

• Task: image text to text
• Library: gguf
• Parameters: 9B
• Weights: about 4.5 GB at 4-bit (estimate)
• Formats: GGUF
• Downloads in the last 30 days: 1,346
• Likes: 13
• License: apache-2.0
• Based on: 169Pi/Alpie_seed
• Tags: gguf, alpie, 169pi, quantized, local-inference, vision-language, multimodal, reasoning, coding, llama.cpp, 9b, apache-2.0

Links

• Model: https://huggingface.co/169Pi/Alpie_Seed_GGUF
• Files: https://huggingface.co/169Pi/Alpie_Seed_GGUF/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download 169Pi/Alpie_Seed_GGUF

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