Only with CPU offload

Gemma 3 12B at Q4_K_M needs 8.3 GB but only 7.0 GB is addressable, so about 19% of the layers would stream from system RAM at roughly 60 GB/s. Expect around 14 tokens per second — usable for batch work, painful for chat.

81% on GPU 8K context Q4_K_M · 6.9 GB Gemma Terms of Use Released Mar 2025 Vision

Strong multilingual chat with images, sized for 12–16 GB cards. Gemma 4 12B is the same size and better.

What hardware do I need for Gemma 3 12B? →

Fits instead: Gemma 3 4B (3.3 GB) · Gemma 3 1B (1.2 GB)

The VRAM budget

weights 6.9 GB

Weights 6.9 GB KV cache @ 8K 0.81 GB Runtime overhead 0.6 GB Over budget 1.3 GB past 7.0 GB

Quantisation ladder

Quant	Weights	Total @ 8K	Max context	Tok/s	Quality	Fit
F16	22.7 GB	24.1 GB	—	~2.0	Reference	17.1 GB over
Q8_0	12.1 GB	13.5 GB	—	~4.7	−0.1% ppl	6.5 GB over
Q6_K	9.3 GB	10.7 GB	—	~7.2	−0.4% ppl	3.7 GB over
Q5_K_M	8.1 GB	9.5 GB	—	~9.6	−0.8% ppl	2.5 GB over
Q4_K_M	6.9 GB	8.3 GB	—	~14	−1.9% ppl	1.3 GB over
Q3_K_M	5.6 GB	7.0 GB	8K	28	−5.4% ppl	Fits
Q2_K	4.8 GB	6.2 GB	21K	33	−15% ppl	Long context

Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it. This model interleaves sliding-window layers (1024 tokens, 1 global in 6), which is why its cache barely grows with context.

How to run it

Ollama llama.cpp LM Studio vLLM

terminal

$ llama-server \
    -hf google/gemma-3-12b-it:Q4_K_M \
    -c 8192 -ngl 39

The engine underneath most of the others. Every knob is exposed. More on llama.cpp.

01Download is 6.9 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.

02Close anything else holding VRAM. A browser with hardware acceleration can sit on 1–2 GB.

03If it falls back to CPU silently, drop the context first, then step down a quantisation.

See all models for this rig Compare with another model