Run lmstudio-community/gemma-4-12B-it-QAT-GGUF locally

License: apache-2.0 ⬇ 598,773 ❤ 9
Parameters11.91B
Context262,144

lmstudio-community/gemma-4-12B-it-QAT-GGUF is a large instruction-tuned chat model with 11.91 billion parameters, built on the gemma4 architecture. It is released under the apache-2.0 license and has been downloaded 598,773 times.

To run lmstudio-community/gemma-4-12B-it-QAT-GGUF locally at a 4,096-token context, its quantized versions need between 2.63 GB (BF16, lowest quality) and 8.96 GB (Q4_0, highest quality) of memory, weights plus KV cache and a system margin included.

For most users the best balance is BF16, needing about 2.63 GB. That means lmstudio-community/gemma-4-12B-it-QAT-GGUF fits entirely in the VRAM of a 6 GB GPU or larger, running fully on the GPU.

→ Guide: How much VRAM do you need?

All quantizations

Quant.Bits QualityWeights KVTotal Speed~Verdict
BF16 0.12 Very low 0.16 GB 1.67 GB 2.63 GB 2452.7 t/s Fits in VRAM
Q4_0 4.69 Good 6.5 GB 1.67 GB 8.96 GB 7.7 t/s Offload

KV cache estimated (architecture unavailable). Speed is a rough estimate bounded by memory bandwidth.

Frequently asked questions

How much VRAM do you need to run lmstudio-community/gemma-4-12B-it-QAT-GGUF?

You need about 2.63 GB of VRAM to run lmstudio-community/gemma-4-12B-it-QAT-GGUF entirely on the GPU using the BF16 quantization (at a 4,096-token context). Smaller quantizations lower the requirement at the cost of quality.

Can I run lmstudio-community/gemma-4-12B-it-QAT-GGUF on an 8 GB GPU?

Yes. With 8 GB of VRAM you can run lmstudio-community/gemma-4-12B-it-QAT-GGUF fully on the GPU using BF16 (about 2.63 GB).

Can I run lmstudio-community/gemma-4-12B-it-QAT-GGUF on a 16 GB GPU?

Yes. With 16 GB of VRAM you can run lmstudio-community/gemma-4-12B-it-QAT-GGUF fully on the GPU using Q4_0 (about 8.96 GB).

Can I run lmstudio-community/gemma-4-12B-it-QAT-GGUF on a 24 GB GPU?

Yes. With 24 GB of VRAM you can run lmstudio-community/gemma-4-12B-it-QAT-GGUF fully on the GPU using Q4_0 (about 8.96 GB).

What is the best quantization for lmstudio-community/gemma-4-12B-it-QAT-GGUF?

If memory allows, higher bits-per-weight means better quality. A common sweet spot is a Q4_K_M or Q5_K_M quantization, which keeps most of the quality while roughly halving the memory versus 8-bit. Pick the highest quantization that still fits in your VRAM.