Benchmark result
google/gemma-4-12b on Apple M1 Max — 29.1 tok/s
Measured with LM Studio on November 3, 2026.
Support llm-bench.io
llm-bench.io is free. If it helped you, I'd appreciate your support to cover the infrastructure costs.
What the model built
Open full screen →The coding scenario asks for a playable game in a single HTML file. This is exactly what the model returned, unedited; the frame only adds a line that tells this page how tall it is.
How this run compares
3rd fastest of 5 runs of this model on Apple M1 Max · median 29.1 tok/s.
sort
- 129.6tok/s68.9LM Studio · 4bit
- 229.5tok/s61.1LM Studio · 4bit
- 329.1tok/s64.5LM Studio · 4bitthis run
- 426.1tok/s62.9LM Studio · 4bit
- 522.7tok/s65.3LM Studio · Q4_K_M
Reproduce this run
toolLM Studio
modelgoogle/gemma-4-12b (4bit)
context65,536
clientv0.4.98+102
Set those in LM Studio, then:
llm-benchmark benchmark --model "google/gemma-4-12b" --tool "LM Studio"The client prompts for context length and thinking mode, and for the KV cache dtype on Unsloth Studio. Sampling is left at the model default — the values above are what the tool reported using, not overrides the client sent.