Skip to content
localmodel.run

Text model · gemma

Gemma 3 27B requirements

gemma family · 27B params · released Mar 2025 · 32.8M Ollama pulls · LMArena Elo 1366. Minimum to run at Q4_K_M: Nvidia GeForce RTX 4090 (24GB).

LicenseGemma· Conditional↓ 1M/mo♥ 2Kon HuggingFace
Q4_K_M
16.55 GB
Q8_0
28.71 GB
Total @ Q4 (4k)
~18.6 GB
Context
128 k

Quantization sizes

GGUF quantson disk
QuantizationSize on disk
Q2_K11.3 GB est
Q3_K_M13.2 GB est
Q4_K_M (default)16.55 GB
Q5_K_M19.2 GB est
Q6_K22.1 GB est
Q8_028.71 GB
FP1654 GB est

Lower quant = smaller and faster, slightly lower quality. Q4_K_M is the common default.

Run it

Ollama
$ ollama run gemma3:27b
llama.cpp
$ llama-cli -hf bartowski/google_gemma-3-27b-it-GGUF:Q4_K_M
LM Studio
$ lms get bartowski/google_gemma-3-27b-it-GGUF

Which devices can run Gemma 3 27B?

FAQ

How much VRAM or RAM does Gemma 3 27B need?

At Q4_K_M, Gemma 3 27B needs about 18.6 GB (weights ~16.55 GB + KV cache + overhead) at a 4k context. At Q8_0 budget ~30.7 GB.

Can Gemma 3 27B run on a laptop?

Gemma 3 27B is large; you need a 24 GB+ GPU or a 32-48 GB Mac at Q4_K_M.

Can I use Gemma 3 27B commercially?

Conditionally. Gemma Terms of Use apply to commercial use.

Released March 12, 2025. Multimodal (text + image), 140+ languages. Ollama shows 17GB default; bartowski HF (google_gemma-3-27b-it-GGUF) gives precise 16.55GB Q4_K_M and 28.71GB Q8_0.

Sources

Memory figures are estimates. See methodology.