GraySoft
Projects Models Compare Cloud benchmarks FAQ Download guIDE →
Model Intelligence Sheet

Serpen/Minimax_M3_MMPROJ_GGUF overview

These are mmproj files for running Minimax M3 with multi modal contexts in llama.cpp. Requires this pr: https://github.com/ggml org/llama.cpp/pull/25113. It wi…

ggufbase_model:MiniMaxAI/MiniMax-M3base_model:quantized:MiniMaxAI/MiniMax-M3region:us

Runs locally from ~1.62 GB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).

Downloads
428
Likes
0
Pipeline
Author

Repository Files & Downloads

1 GGUF files detected
Direct downloads for local inference
FileTypeQuantizationSizeLink
M3-mmproj.ggufGGUFGGUF1.62 GBDownload

Model Details

Model IDSerpen/Minimax_M3_MMPROJ_GGUF
AuthorSerpen
Pipeline
License
Base modelMiniMaxAI/MiniMax-M3
Last modified2026-07-01T22:03:53.000Z

Model README

---

base_model:

  • MiniMaxAI/MiniMax-M3

---

These are mmproj files for running Minimax-M3 with multi modal contexts in llama.cpp. Requires this pr: https://github.com/ggml-org/llama.cpp/pull/25113. It will NOT load with Unsloth's or the non vision MSA PR. Furthermore the linked PR depends on Minimax Sparse Attention, so Unsloth's GGUFs will NOT load on it.

You will need to convert or download GGUFs with the Indexer tensors preserved: https://huggingface.co/Serpen/Minimax-M3-MSA-GGUF. Both Vision and MSA are fully functional, at low ctx decoding speed is around the same as Unsloth's PR for now, but MSA starts pulling ahead at longer context sizes.

What MSA helps with then you may ask?

ELI5: the model was trained to only ever show 2048 tokens of ctx to any single indexer head. (Chosen individually for each of the 4 indexer heads for every sparse layer). This is what the model was trained on. Deviating from this will cause long context recall issues. Unsloth's PR instead shows the full context to every single one of these heads. This dilutes attention, and pushes the model out of distribution.

Run Serpen/Minimax_M3_MMPROJ_GGUF with guIDE

Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.

Download guIDE → · Browse 524k+ models · Compare models

Source: Hugging Face · Compare models