CoolFace
Modelpublic

EnlistedGhost/Pixtral-12B-Ollama-GGUF

sourceHugging Faceapache-2.0updated 1mo agoView on Hugging Face
2likes2.9kdownloads
Model Card

Model - Pixtral 12B (GGUF) (Ollama Patched)

Description: <br /> This is an Ollama patched version of Pixtral-12B (GGUF) with working projector (mmproj) files. <br /> No modifications, edits, or configuration is required to use this model with Ollama, it works natively! <br /> Both Vision and Text work with Ollama. (^.^)

IMPORTANT NOTICE as of (GMT-8) 01:04 November 19th, 2025: <br /> After significant amounts of self-testing of the supplied Quantized GGUF files made available within this release <b /> I have been pleasantly surprised to find that so far ALL of the files have the AI performing quite admirably despite <b /> Huggingface's systems here only merging in the Q8_0 (8-bit quantized) Vision Projector - with no automated option to <b /> choose a higher quality beyond assuming that you already know how to use Ollama's "create" command and so on.

I will be updating this release with both:

  • —Instructions within this model page and linked documentation on how to easily "create" a customized/tailored <br /> Multi-Modal model that you may speak/inference with utilizing Ollama (or llama.cpp as well).
  • —Automated scripts via bash (Linux users) and command-line (Windows users) options that will allow anyone to <br /> simply download the small script file and then merely run (click on) the script which will interactively guide you <b /> through a user-friendly setup process - including merging the higher quality Vision Projector options.

In addition to the above mentioned updates:

I am currently releasing GGUF Ollama-ready (Natively Ollama compatible) alternate versions of <b /> Pixtral-12B. In case some of you are not aware of this fact: there are currently two other Officially released versions <b /> of Pixtral-12B. These are supplied directly from MistralAI themselves and include the Pixtral-12B-2409 official release and <b /> the the Pixtral-12B-Base-2409 models that, unlike the release that these files were converted and quantized from, ARE NOT a community release. <br /> These two other versions differ in quality and capability, although slightly, from the originating "mistral-community" release <b /> that this GGUF release (and MOST other releases or derivatives) is based upon. This is because MistralAI's two releases with the "2409" label <b /> are not user-friendly to setup or use. Thankfully I have humbly managed to have them successfully loaded, converted to GGUF and now <b /> am in the process of creating quality Quantized GGUF versions of specifically for use with Ollama/Llama.cpp.

Be sure to check back occasionally as I expect to make these more rarely seen Pixtral-12B releases from MistralAI made available within the next couple days! (Rarely seen - to clarify what I meant was that they are extremely uncommong to find GGUF versions of, and not many versions even that are not GGUF)

Happy Inferencing! <br /> -- Jon Z (EnlistedGhost)

Model Updates (As of: November 19th, 2025)

Recently finished updates:

  • —Partially Updated: ModelCard <br /> (this page)

Currently in-progress updates:

  • —Update: ModelCard <br /> (this page needs more setup guides and information...)

How to run this Model using Ollama

You can run this model by using the "ollama run" command.<br /> Simply copy & paste one of the commands from the list below into<br /> your console, terminal or power-shell window. | Quant Type | File Size | Command | |:-----------|:----------|:--------| | Q2KS | 4.62 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q2KS | | Q2K | 4.79 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q2K | | IQ2M | 4.44 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:IQ2M | | Q2KM | 4.95 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q2KM | | Q2KL | 5.58 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q2KL | | IQ3S | 5.56 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:IQ3S | | Q3KS | 5.53 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q3KS | | IQ3M | 5.72 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:IQ3M | | Q3KM | 6.08 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q3KM | | Q3KL | 6.56 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q3KL | | Q3KXL | 6.72 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q3KXL | | IQ4XS | 6.8 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:IQ4XS | | Q4KS | 7.12 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q4KS | | Q4KM | 7.48 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q4KM | | Q4KXL |8.27 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q4KXL | | Q5KS | 8.52 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q5KS | | Q5KM | 8.73 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q5KM | | Q5KXL | 9.52 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q5KXL | | Q6K | 10.1 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q6K | | Q6KM | 10.2 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q6KM | | Q6KL | 10.8 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q6KL | | Q80 | 13.7 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:Q80 | | F16 | 24.5 GB | ollama run hf.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF:F16

mmproj (Vision Projector) Files | Quant Type | File Size | Command | |:-----------|:----------|:--------| | Q80 | 465 MB | [[mmproj Vision Pixtral-12B rojector:Q80](https://huggingface.co/EnlistedGhost/Pixtral-12B-Ollama-GGUF/resolve/main/mmproj-Pixtral-12B-Q8_0.gguf?download=true)]| | F16 | 870 MB | [mmproj Vision Pixtral-12B rojector:F16]| | F32 | 1.74 GB | [mmproj Vision Pixtral-12B rojector:F32]|

Intended Use

Same as original:

Out-of-Scope Use

Same as original:

Bias, Risks, and Limitations

Same as original:

Training Details

Training sets and data are from:

Evaluation

  • —This model has NOT been evaluated in any form, scope or type of method.
  • —!!! USE AT YOUR OWN RISK !!!
  • —!!! NO WARRANTY IS PROVIDED OF ANY KIND !!!

Citation (Original Paper)

[MistalAI Pixtral-12B Original Paper]

Detailed Release Information

Attributions (Credits)

A big thank-you is extended to the below credited sources!<br /> These contributions are what made this release possible!

Important Notice: This is NOT a copy/paste release, <br /> I have created unique Quantized files that were then altered further <br /> to properly work with Ollama software. <br /> This resulted in the first publicly available <br /> Pixtral-12B model that natively runs on Ollama.

Model Card Authors and Contact

[EnlistedGhost]