Prefill and token generation speed
Gemma 4 12B on Mac mini M4 24GB: chat speed
Try Gemma 4 12B on a Mac mini M4 24GB. Watch the initial wait and the response stream to judge whether the pace suits document summaries or short conversations.
Benchmark-based estimate · 4K input · one user
Last updated
11.6 ~ 12.9 tok/sDecode speed
Model fully loaded (8GB / 20GB)
First token (TTFT)
12.8 ~ 31.1 sContext 4,096 input tokens included
Prefill speed
132 ~ 323 tok/s12.7 to 31.0 seconds (Estimation based on public benchmarks)
Total time (E2E)
Approximately 21.8 to 41.1 seconds117 output tokens included
Press Play to process the input and generate the answer.
Ready
Change device or model
120 GB/s
11.95B
Expand settings Settings4-bit · 4K · Baseline decodeCollapse
QuantizationNeed 8GB
Input context4K tokens
Baseline
Baseline
Sample question:
117 output tokensReading this result
Not all 24GB is available to the model. Compare its requirements with memory left after the operating system and other apps. Longer documents also change waiting time and memory use.
- Required / available memory
- 8.0GB / 20GB
- Runtime
- MLX 4-bit
Compare this model side by side
If you are considering 32GB, compare the same Gemma model first. Keeping the model fixed helps distinguish extra capacity from a change in workload.
Check configuration
Mac mini M4 24GB · Memory 24GB
Gemma 4 12B Q4_K_M configuration. Check that the retailer lists the same memory option.
What changes with a longer document?Expand input length, waiting time, speed and memory
Detailed analysis and all 60 hardware configurations
Popular configurations