This repository provides a 6-bit MLX-quantized GGUF version of Google's Gemma 4 31B instruction-tuned (it) model, optimized for local inference on Apple devices. It includes an updated chat template with improved tool-calling and reasoning capabilities. The model is part of the LM Studio community collection and suitable for developers seeking high-performance local AI.
In the Quantised & converted weights space, Gemma 4 31B It takes a focused approach. It focuses on running a large 31B parameter instruction-tuned LLM efficiently on Apple silicon hardware using MLX. It is built as an open-source project for developers. Gemma 4 31B It is open source under the MIT license. It ships for the web, the command line, and API.
lmstudio-community builds and maintains Gemma 4 31B It, and it first shipped in 2024. The project is developed in the open on GitHub with 5.3k stars and 541 commits in the last 90 days. The category is crowded — PulseGate's index counts 25 comparable projects. Key capabilities include instruction tuning, MLX quantization, and tool calling support.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do