Created bymario-alka.dePowered bynoob2claw.detricoma.de
Google

gemma-3n-E4B-it8B

Relabel 2026-09-23: Laeufe waren faelschlich als gemma-4-E4B-it erfasst (GGUF-Metadata = Gemma 3n E4B)

Dense8B
Position in the field

Best values compared

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

Generation273,1 tok/s
Min 0,1Ø 159,1Max 717,6
Ueber dem Durchschnitt · 361 Systeme im Feld
Prefill4.998 tok/s
Min 39Ø 5.819Max 52.503
Unter dem Durchschnitt · 361 Systeme im Feld
Time to First Token617 ms
Min 156Ø 18.353Max 354.928
Ueber dem Durchschnitt · 361 Systeme im Feld
Performance profile

Throughput by hardware & engine

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

GPUby graphics card

34625917386,40,05681.7652.9634.160Prefill (tok/s)Generation (tok/s)AMD Radeon RX 7900 XTX - 273,1 tok/s Generation, 3.529 tok/s Prefill, TTFT 2.950 ms (9 Laufe)AMD Radeon RX 7900 XTXCPU-only - 46,8 tok/s Generation, 1.199 tok/s Prefill, TTFT 1.907 ms (2 Laufe)CPU-only
AMD Radeon RX 7900 XTX 273,1 tok/sCPU-only 46,8 tok/s

CPUby processor

34726017486,80,0651.5833.1014.619Prefill (tok/s)Generation (tok/s)AMD Ryzen Threadripper PRO 3955WX 16-Cores - 273,1 tok/s Generation, 3.847 tok/s Prefill, TTFT 2.646 ms (6 Laufe)AMD Ryzen Threadrippe...Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz - 220,8 tok/s Generation, 2.894 tok/s Prefill, TTFT 3.558 ms (3 Laufe)Intel(R) Xeon(R) CPU ...Apple M6 (unified CPU/GPU SoC, ARM64, AdvSIMD) - 46,8 tok/s Generation, 1.560 tok/s Prefill, TTFT 1.426 ms (1 Lauf)Apple M6 (unified CPU...Apple M6 (unified CPU/GPU SoC) - 39,4 tok/s Generation, 838 tok/s Prefill, TTFT 2.388 ms (1 Lauf)Apple M6 (unified CPU...
AMD Ryzen Threadripper PRO 3955WX 16-Cores 273,1 tok/sIntel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz 220,8 tok/sApple M6 (unified CPU/GPU SoC, ARM64, AdvSIMD) 46,8 tok/sApple M6 (unified CPU/GPU SoC) 39,4 tok/s

ENGby engine

34726017486,80,01421.5032.8644.225Prefill (tok/s)Generation (tok/s)llama.cpp - 273,1 tok/s Generation, 3.529 tok/s Prefill, TTFT 2.950 ms (9 Laufe)llama.cpplmstudio-mlx - 46,8 tok/s Generation, 1.560 tok/s Prefill, TTFT 1.426 ms (1 Lauf)lmstudio-mlxOllama - 39,4 tok/s Generation, 838 tok/s Prefill, TTFT 2.388 ms (1 Lauf)Ollama
llama.cpp 273,1 tok/slmstudio-mlx 46,8 tok/sOllama 39,4 tok/s
Throughput & latency

Performance benchmark

Metric:
GPUs
#Model / MakerMetricsParallelGPU / CPU / RAMRuntime
1gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
273,10 tok/s TG
Prefill 4.691 · TTFT 4.615 ms
10×AMD Radeon RX 7900 XTXAMD Ryzen Threadripper PRO 3955WX 16-Cores · 63 GB RAMllama.cppgodclawQ4_K_M
2gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
271,11 tok/s TG
Prefill 4.998 · TTFT 4.486 ms
10×AMD Radeon RX 7900 XTXAMD Ryzen Threadripper PRO 3955WX 16-Cores · 63 GB RAMllama.cppgodclawQ8_0
3gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
220,77 tok/s TG
Prefill 3.758 · TTFT 6.047 ms
10×2x AMD Radeon RX 7900 XTX i4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAMllama.cppgodclawQ4_K_M
4gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
189,79 tok/s TG
Prefill 3.528 · TTFT 2.818 ms
5×AMD Radeon RX 7900 XTXAMD Ryzen Threadripper PRO 3955WX 16-Cores · 63 GB RAMllama.cppgodclawQ4_K_M
5gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
188,89 tok/s TG
Prefill 3.739 · TTFT 2.671 ms
5×AMD Radeon RX 7900 XTXAMD Ryzen Threadripper PRO 3955WX 16-Cores · 63 GB RAMllama.cppgodclawQ8_0
6gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
116,57 tok/s TG
Prefill 2.698 · TTFT 3.741 ms
5×2x AMD Radeon RX 7900 XTX i4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAMllama.cppgodclawQ4_K_M
7gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
77,64 tok/s TG
Prefill 2.929 · TTFT 673 ms
1×AMD Radeon RX 7900 XTXAMD Ryzen Threadripper PRO 3955WX 16-Cores · 63 GB RAMllama.cppgodclawQ4_K_M
8gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
66,77 tok/s TG
Prefill 3.195 · TTFT 617 ms
1×AMD Radeon RX 7900 XTXAMD Ryzen Threadripper PRO 3955WX 16-Cores · 63 GB RAMllama.cppgodclawQ8_0
9gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
48,95 tok/s TG
Prefill 2.226 · TTFT 887 ms
1×2x AMD Radeon RX 7900 XTX i4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz · 504 GB RAMllama.cppgodclawQ4_K_M
10gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
46,82 tok/s TG
Prefill 1.560 · TTFT 1.426 ms
1×Keine GPU (CPU-only)Apple M6 (unified CPU/GPU SoC, ARM64, AdvSIMD) · 24 GB RAMlmstudio-mlxlmstudioMLX 4bit
11gemma-3n-E4B-it8BGoogle Performance benchmarkTimebench 3 - Kombi (Prefill + Generation)
39,36 tok/s TG
Prefill 838 · TTFT 2.388 ms
1×Keine GPU (CPU-only)Apple M6 (unified CPU/GPU SoC) · 24 GB RAMOllamaopenclaw_cliQ4_K_M
Agent & chat rating

Harness benchmark

Noch keine Harnessbenchmarks fuer dieses Modell.