Google

Gemma-4-26B-A4B26B

Name: Gemma-4-26B-A4B
Author: Google

Benchmark profile and published results.

MoE26B

Position in the field

Best values compared

Best metrics of this model against the minimum, average and maximum of all published systems.

Generation47,1 tok/s

Min 11,4Ø 148,9Max 903,3

Unter dem Durchschnitt · 82 Systeme im Feld

Prefill4.378 tok/s

Min 503Ø 3.550Max 7.470

Ueber dem Durchschnitt · 82 Systeme im Feld

Time to First Token451 ms

Min 105Ø 8.759Max 66.475

Ueber dem Durchschnitt · 82 Systeme im Feld

Performance profile

Each bubble is a GPU, CPU or engine – position shows prompt processing (X) and output speed (Y), bubble size the number of runs.

NVIDIA GB10 47,1 tok/s

NVIDIA Grace 47,1 tok/s

vLLM 47,1 tok/s

Throughput & latency

Metric:

	#	Model / Maker	Metrics	Parallel	GPU / CPU / RAM	Runtime
	1	Gemma-4-26B-A4B26BNVFP4Google Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)	47,06 tok/s TG Prefill 4.378 · TTFT 451 ms	1×	NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM	vLLMgodclaw	Details →

Agent & chat rating

Noch keine Harnessbenchmarks fuer dieses Modell.