No — not on this device
Gemma 3 27B at Q4_K_M needs 17.0 GB against 10.7 GB usable, and the shortfall of 6.3 GB is more than 128 GB of system RAM can cover at a tolerable speed. A smaller sibling or a lower quantisation is the honest answer here.
The 2025 single-GPU generalist with vision. Its Gemma-licence terms are the reason to prefer Gemma 4 now.
What hardware do I need for Gemma 3 27B? →
Fits instead: Gemma 3 12B (8.3 GB) · Gemma 3 4B (3.3 GB)
The VRAM budget
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| Q8_0 | 27.1 GB | 28.7 GB | — | ~1.4 | −0.1% ppl | 18.0 GB over |
| Q6_K | 20.9 GB | 22.6 GB | — | ~1.8 | −0.4% ppl | 11.9 GB over |
| Q5_K_M | 18.1 GB | 19.7 GB | — | ~2.1 | −0.8% ppl | 9.0 GB over |
| Q4_K_M | 15.4 GB | 17.0 GB | — | ~2.5 | −1.9% ppl | 6.3 GB over |
| Q3_K_M | 12.5 GB | 14.1 GB | — | ~3.0 | −5.4% ppl | 3.4 GB over |
| Q2_K | 10.7 GB | 12.3 GB | — | ~3.6 | −15% ppl | 1.6 GB over |
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
$ pip install mlx-lm $ mlx_lm.generate --model mlx-community/gemma-3-27b-it-4bit \ --max-tokens 512 --prompt "Hello"
Apple's own array framework. The fastest path on Apple Silicon. More on MLX.