Model benchmarks

gemma4:12b local LLM performance

As of July 2026, gemma4:12b runs at up to 46.3 tok/s for local inference (best of 2 community benchmark runs across 0 GPUs).

OllamaQ4_K_M

Model size

11.9B

Peak speed

46.3 tok/s

Average speed

46.3 tok/s

Min memory

7.8 GB

Max context

5,347 tokens

Best quality

81.3

Benchmark runs

2

GPUs tested

0

Performance by hardware and tool

Every hardware/tool/quantization combination gemma4:12b has been benchmarked on, ranked by peak token generation speed. Last updated July 2026.

HardwareToolQuantBest tok/sAvg tok/sMemoryContextQualityRuns
CPU onlyOllamaQ4_K_M46.3 tok/s46.3 tok/s7.8 GB5,347 tokens68.52

Frequently asked questions

How fast is gemma4:12b for local inference?
Across 2 community benchmark runs, gemma4:12b reaches up to 46.3 tok/s and averages 46.3 tok/s.
How much memory does gemma4:12b need?
The leanest observed configuration used about 7.8 GB of memory (quantizations tested: Q4_K_M).
Which tools have been used to run gemma4:12b?
Benchmarks were submitted using Ollama. Results are community-contributed and updated as new runs arrive.