Gemma 4 E2B
109 records across 14 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M4 Max | Gemma 4 E2B | oQ4e | oMLX | 379.0 | 7935.0 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | Gemma 4 E2B | oQ4e | oMLX | 376.3 | 7923.0 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | Gemma 4 E2B | oQ4e | oMLX | 347.2 | 7881.0 | – | – | 4096 | 2026-10-11 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 201.9 | 11106.0 | – | – | 4096 | 2026-10-11 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 182.8 | 10696.0 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Max | Gemma 4 E2B | 4bit | oMLX | 180.6 | 3400.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 172.7 | 9855.0 | – | – | 16384 | 2026-10-11 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 146.3 | 9789.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | Gemma 4 E2B | oQ4e | oMLX | 144.6 | 7956.0 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 141.3 | 9249.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 135.9 | 7929.0 | – | – | 1024 | 2026-10-09 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 134.5 | 8580.0 | – | – | 32768 | 2026-10-11 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 130.6 | 10894.0 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 130.5 | 7858.0 | – | – | 1024 | 2026-10-09 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 126.7 | 10542.0 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 125.9 | 10452.0 | – | – | 4096 | 2026-10-09 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 125.7 | 10439.0 | – | – | 4096 | 2026-10-09 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 123.5 | 10141.0 | – | – | 8192 | 2026-10-09 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 121.5 | 10203.0 | – | – | 8192 | 2026-10-09 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 119.1 | 9865.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 118.7 | 9720.0 | – | – | 16384 | 2026-10-09 | community | contradiction | source |
| M1 Ultra | Gemma 4 E2B | oQ6e | oMLX | 116.9 | 2098.0 | – | – | 1024 | 2026-10-10 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 107.1 | 8604.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 103.6 | 8281.0 | – | – | 32768 | 2026-10-09 | community | contradiction | source |
| M2 Max | Gemma 4 E2B | oQ8 | oMLX | 101.7 | 2936.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M2 Max | Gemma 4 E2B | oQ8 | oMLX | 101.7 | 2936.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M2 Max | Gemma 4 E2B | oQ8 | oMLX | 100.9 | 2728.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| M2 Max | Gemma 4 E2B | oQ8 | oMLX | 100.2 | 4988.0 | – | – | 4096 | 2026-10-11 | community | – | source |
| M2 Max | Gemma 4 E2B | oQ8 | oMLX | 100.2 | 4988.0 | – | – | 4096 | 2026-10-11 | community | – | source |
| M5 Pro | Gemma 4 E2B | nvfp4 | oMLX | 94.4 | 3980.0 | – | – | 4096 | 2026-10-07 | community | – | source |
| M4 Max | Gemma 4 E2B | oQ4e | oMLX | 93.4 | 5325.0 | – | – | 4096 | 2026-10-11 | community | contradiction outlier | source |
| M4 Max | Gemma 4 E2B | oQ4e | oMLX | 90.2 | 8029.0 | – | – | 4096 | 2026-10-11 | community | contradiction outlier | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 89.2 | 1394.0 | – | – | 1024 | 2026-10-13 | community | – | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 88.8 | 3152.0 | – | – | 4096 | 2026-10-13 | community | – | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 84.9 | 3509.0 | – | – | 4096 | 2026-10-08 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 84.8 | 6228.0 | – | – | 65536 | 2026-10-09 | community | – | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 82.5 | 2384.0 | – | – | 4096 | 2026-10-17 | community | – | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 80.0 | 2295.0 | – | – | 1024 | 2026-10-17 | community | – | source |
| M1 Max | Gemma 4 E2B | 8bit | oMLX | 78.9 | 1560.0 | – | – | 1024 | 2026-10-09 | community | – | source |
| M1 Max | Gemma 4 E2B | 8bit | oMLX | 76.6 | 2661.0 | – | – | 4096 | 2026-10-09 | community | – | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 74.0 | 2389.0 | – | – | 16384 | 2026-10-08 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 65.3 | 2347.0 | – | – | 4096 | 2026-10-07 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 63.5 | 2343.0 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 62.9 | 2018.0 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 62.9 | 2018.0 | – | – | 4096 | 2026-10-12 | community | – | source |
| M5 Pro | Gemma 4 E2B | 4bit | oMLX | 61.9 | 4153.0 | – | – | 131072 | 2026-10-09 | community | – | source |
| M3 | Gemma 4 E2B | 4bit | oMLX | 61.5 | 1436.0 | – | – | 1024 | 2026-10-13 | community | – | source |
| M6 | Gemma 4 E2B | 5bit | oMLX | 59.6 | 4962.0 | – | – | 32768 | 2026-10-08 | community | – | source |
| M3 | Gemma 4 E2B | 4bit | oMLX | 58.1 | 1900.0 | – | – | 4096 | 2026-10-13 | community | – | source |
| M1 Pro | Gemma 4 E2B | 4bit | oMLX | 57.7 | 1587.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 57.1 | 2048.0 | – | – | 16384 | 2026-10-12 | community | – | source |
| M1 Pro | Gemma 4 E2B | 4bit | oMLX | 56.1 | 1873.0 | – | – | 4096 | 2026-10-15 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 56.0 | 2038.0 | – | – | 16384 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 4bit | oMLX | 56.0 | 2038.0 | – | – | 16384 | 2026-10-12 | community | – | source |
| M1 Pro | Gemma 4 E2B | 4bit | oMLX | 54.0 | 1380.0 | – | – | 1024 | 2026-10-06 | community | – | source |
| M1 Pro | Gemma 4 E2B | 4bit | oMLX | 53.8 | 1605.0 | – | – | 4096 | 2026-10-06 | community | – | source |
| M2 | Gemma 4 E2B | 4bit | oMLX | 50.9 | 1260.0 | – | – | 4096 | 2026-10-07 | community | – | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 50.8 | 1496.0 | – | – | 65536 | 2026-10-08 | community | – | source |
| M6 | Gemma 4 E2B | 5bit | oMLX | 47.9 | 4486.0 | – | – | 65536 | 2026-10-08 | community | – | source |
| M2 | Gemma 4 E2B | 4bit | oMLX | 43.6 | 1295.0 | – | – | 16384 | 2026-10-07 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 39.0 | 2099.0 | – | – | 1024 | 2026-10-17 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 39.0 | 2099.0 | – | – | 1024 | 2026-10-17 | community | – | source |
| M1 | Gemma 4 E2B | 4bit | oMLX | 38.6 | 984.6 | – | – | 1024 | 2026-10-09 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 38.5 | 2318.0 | – | – | 4096 | 2026-10-17 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 38.5 | 2318.0 | – | – | 4096 | 2026-10-17 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 38.4 | 2227.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 38.0 | 2345.0 | – | – | 4096 | 2026-10-15 | community | – | source |
| M1 | Gemma 4 E2B | 4bit | oMLX | 37.9 | 987.6 | – | – | 1024 | 2026-10-16 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 36.9 | 2194.0 | – | – | 1024 | 2026-10-10 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 36.1 | 2261.0 | – | – | 4096 | 2026-10-10 | community | – | source |
| M1 | Gemma 4 E2B | 4bit | oMLX | 35.8 | 987.7 | – | – | 4096 | 2026-10-16 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 35.1 | 1950.0 | – | – | 1024 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 35.1 | 1950.0 | – | – | 1024 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.9 | 2207.0 | – | – | 1024 | 2026-10-14 | community | contradiction | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.9 | 2222.0 | – | – | 1024 | 2026-10-14 | community | contradiction | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.7 | 2219.0 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.7 | 2219.0 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.6 | 2259.0 | – | – | 4096 | 2026-10-14 | community | contradiction | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.5 | 2222.0 | – | – | 1024 | 2026-10-14 | community | contradiction | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.2 | 2291.0 | – | – | 4096 | 2026-10-14 | community | contradiction | source |
| M6 | Gemma 4 E2B | 5bit | oMLX | 34.1 | 2855.0 | – | – | 131072 | 2026-10-08 | community | – | source |
| M4 | Gemma 4 E2B | 8bit | oMLX | 34.0 | 2306.0 | – | – | 4096 | 2026-10-14 | community | contradiction | source |
| M1 Max | Gemma 4 E2B | 4bit | oMLX | 32.4 | 1269.0 | – | – | 131072 | 2026-10-08 | community | – | source |
| M6 | Gemma 4 E2B | 5bit | oMLX | 27.3 | 2083.0 | – | – | 199680 | 2026-10-08 | community | – | source |
| M1 | Gemma 4 E2B | 8bit | oMLX | 15.7 | 664.2 | – | – | 32768 | 2026-10-14 | community | – | source |
| OPPO Reno7 (Dimensity 900) | Gemma 4 E2B | Q4_0 | llama.cpp (CPU) | 4.5 | 16.0 | – | – | 8192 | 2026-10-02 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 30314.2 | – | – | 2048 | 2026-10-13 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 38015.2 | – | – | 2048 | 2026-10-13 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 40572.3 | – | – | 2048 | 2026-10-13 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 30697.3 | – | – | 4096 | 2026-10-13 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 35937.9 | – | – | 4096 | 2026-10-13 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 25433.3 | – | – | 1024 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 35480.8 | – | – | 1024 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 37836.3 | – | – | 1024 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 39380.0 | – | – | 1024 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 12762.8 | – | – | 512 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 27009.7 | – | – | 512 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 31879.1 | – | – | 512 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 36008.9 | – | – | 512 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 2489.8 | – | – | 128 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 8692.8 | – | – | 128 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 13812.5 | – | – | 128 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 21129.3 | – | – | 128 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 30383.4 | – | – | 4096 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 32728.3 | – | – | 4096 | 2026-10 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 35480.8 | – | – | – | 2026-10-08 | community | – | source |
| A100-SXM4-80GB | Gemma 4 E2B | BF16 | vLLM | – | 67488.9 | – | – | – | 2026-10-08 | community | – | source |
| X2-90 | Gemma 4 E2B | Q4_0 | llama.cpp | – | 881 | – | – | 512 | 2026-10-14 | sourced | – | source |
| X2-90 | Gemma 4 E2B | Q4_0 | llama.cpp | – | 1028 | – | – | 512 | 2026-10-14 | sourced | – | source |