gemma-4-31B-it
209 records across 15 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M5 Ultra | gemma-4-31B-it | oQ8e | oMLX | 59.6 | 923.8 | – | – | 16384 | 2026-10-15 | community | – | source |
| M5 Ultra | gemma-4-31B-it | oQ8e | oMLX | 52.0 | 975.7 | – | – | 4096 | 2026-10-15 | community | – | source |
| M5 Ultra | gemma-4-31B-it | oQ8e | oMLX | 50.6 | 960.9 | – | – | 8192 | 2026-10-15 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 48.3 | 703.9 | – | – | 1024 | 2026-10-07 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 46.7 | 204.4 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 46.7 | 204.4 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 46.7 | 204.4 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 46.7 | 204.4 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 46.7 | 204.4 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 46.7 | 204.4 | – | – | 4096 | 2026-10-11 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 45.8 | 601.3 | – | – | 16384 | 2026-10-07 | community | – | source |
| M5 Ultra | gemma-4-31B-it | oQ8e | oMLX | 45.5 | 845.0 | – | – | 32768 | 2026-10-15 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 45.2 | 727.8 | – | – | 4096 | 2026-10-07 | community | – | source |
| M5 Ultra | gemma-4-31B-it | oQ8e | oMLX | 44.5 | 911.1 | – | – | 1024 | 2026-10-15 | community | – | source |
| M3 Ultra | gemma-4-31B-it | oQ6e | oMLX | 42.4 | 254.9 | – | – | 4096 | 2026-10-10 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 41.7 | 657.2 | – | – | 16384 | 2026-10-17 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 39.8 | 693.0 | – | – | 4096 | 2026-10-17 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 39.1 | 193.4 | – | – | 16384 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 39.1 | 193.4 | – | – | 16384 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 39.1 | 193.4 | – | – | 16384 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 39.1 | 193.4 | – | – | 16384 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 39.1 | 193.4 | – | – | 16384 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 39.1 | 193.4 | – | – | 16384 | 2026-10-11 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 38.3 | 665.4 | – | – | 8192 | 2026-10-07 | community | – | source |
| M3 Ultra | gemma-4-31B-it | oQ8e | oMLX | 38.0 | 344.3 | – | – | 1024 | 2026-10-09 | community | – | source |
| M3 Ultra | gemma-4-31B-it | oQ4e | oMLX | 37.7 | 259.7 | – | – | 4096 | 2026-10-10 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 37.6 | 644.9 | – | – | 16384 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 35.2 | 676.9 | – | – | 4096 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 35.2 | 676.9 | – | – | 4096 | 2026-10-12 | community | contradiction | source |
| M3 Ultra | gemma-4-31B-it | oQ6e | oMLX | 35.1 | 252.5 | – | – | 1024 | 2026-10-10 | community | – | source |
| M5 Ultra | gemma-4-31B-it | oQ8e | oMLX | 34.9 | 723.8 | – | – | 65536 | 2026-10-15 | community | – | source |
| M3 Ultra | gemma-4-31B-it | oQ8e | oMLX | 34.1 | 346.3 | – | – | 4096 | 2026-10-09 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 33.1 | 167.8 | – | – | 1024 | 2026-10-12 | community | – | source |
| M3 Ultra | gemma-4-31B-it | oQ4e | oMLX | 32.4 | 246.6 | – | – | 16384 | 2026-10-10 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 32.3 | 592.3 | – | – | 4096 | 2026-10-09 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 32.0 | 491.4 | – | – | 65536 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 4bit | oMLX | 31.9 | 208.5 | – | – | 1024 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 31.6 | 208.0 | – | – | 1024 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 31.6 | 208.0 | – | – | 1024 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 30.7 | 193.4 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 30.7 | 193.4 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 30.7 | 193.4 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 30.7 | 193.4 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 30.7 | 193.4 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 30.7 | 193.4 | – | – | 8192 | 2026-10-11 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 30.5 | 623.4 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 30.5 | 205.7 | – | – | 1024 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 30.1 | 575.1 | – | – | 1024 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 29.7 | 476.6 | – | – | 16384 | 2026-10-09 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 29.7 | 688.5 | – | – | 4096 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 29.3 | 468.2 | – | – | 16384 | 2026-10-09 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 28.9 | 651.8 | – | – | 16384 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 28.9 | 651.8 | – | – | 16384 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 28.6 | 507.2 | – | – | 16384 | 2026-10-09 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 28.5 | 207.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 28.5 | 207.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 28.5 | 207.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 28.5 | 207.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 28.5 | 207.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 28.5 | 207.0 | – | – | 1024 | 2026-10-11 | community | – | source |
| Intel Arc B580 Limited Edition | gemma-4-31B-it | UD-Q4_K_XL | llama.cpp (SYCL) | 28.44 | – | – | – | 65536 | 2026-09-25 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 28.4 | 185.1 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 28.4 | 185.1 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 28.4 | 172.7 | – | – | 4096 | 2026-10-12 | community | – | source |
| M5 Ultra | gemma-4-31B-it | 8bit | oMLX | 28.2 | 879.2 | – | – | 1024 | 2026-10-10 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 28.2 | 186.2 | – | – | 4096 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 28.1 | 541.7 | – | – | 1024 | 2026-10-12 | community | – | source |
| M5 Ultra | gemma-4-31B-it | 8bit | oMLX | 28.0 | 888.4 | – | – | 1024 | 2026-10-09 | community | – | source |
| M5 Ultra | gemma-4-31B-it | 8bit | oMLX | 27.8 | 864.8 | – | – | 4096 | 2026-10-10 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 27.6 | 564.9 | – | – | 32768 | 2026-10-07 | community | – | source |
| M5 Ultra | gemma-4-31B-it | 8bit | oMLX | 27.5 | 973.8 | – | – | 4096 | 2026-10-09 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 26.2 | 179.7 | – | – | 32768 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 26.2 | 179.7 | – | – | 32768 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 26.2 | 179.7 | – | – | 32768 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 26.2 | 179.7 | – | – | 32768 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 26.2 | 179.7 | – | – | 32768 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 26.2 | 179.7 | – | – | 32768 | 2026-10-11 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 26.1 | 659.2 | – | – | 4096 | 2026-10-09 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 25.6 | 708.3 | – | – | 1024 | 2026-10-16 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 24.0 | 692.7 | – | – | 4096 | 2026-10-16 | community | contradiction | source |
| Intel Arc B580 Limited Edition | gemma-4-31B-it | UD-Q4_K_XL | llama.cpp (SYCL) | 23.79 | – | – | – | 65536 | 2026-09-25 | community | – | source |
| M4 Max | gemma-4-31B-it | 4bit | oMLX | 23.4 | 207.4 | – | – | 1024 | 2026-10-12 | community | contradiction | source |
| Intel Arc B580 Limited Edition | gemma-4-31B-it | UD-Q4_K_XL | Vulkan | 23.11 | – | – | – | 65536 | 2026-09-25 | community | – | source |
| M4 Max | gemma-4-31B-it | 4bit | oMLX | 23.1 | 188.3 | – | – | 4096 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 23.0 | 205.8 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 23.0 | 205.8 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 23.0 | 205.8 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 23.0 | 205.8 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 23.0 | 205.8 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 23.0 | 205.8 | – | – | 4096 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 22.4 | 197.3 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 22.4 | 197.3 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 22.4 | 197.3 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 22.4 | 197.3 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 22.4 | 197.3 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 22.4 | 197.3 | – | – | 8192 | 2026-10-11 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 21.9 | 627.1 | – | – | 32768 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 4bit | oMLX | 21.9 | 612.4 | – | – | 1024 | 2026-10-13 | community | – | source |
| M5 Max | gemma-4-31B-it | 4bit | oMLX | 21.8 | 587.3 | – | – | 1024 | 2026-10-10 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 21.5 | 193.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 21.5 | 193.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 21.5 | 193.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 21.5 | 193.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 21.5 | 193.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 21.5 | 193.0 | – | – | 16384 | 2026-10-11 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 4bit | oMLX | 21.4 | 604.0 | – | – | 4096 | 2026-10-10 | community | – | source |
| M5 Max | gemma-4-31B-it | 4bit | oMLX | 21.3 | 589.5 | – | – | 4096 | 2026-10-13 | community | – | source |
| Intel Arc B580 Limited Edition | gemma-4-31B-it | UD-Q4_K_XL | Vulkan | 20.87 | – | – | – | 65536 | 2026-09-25 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 20.7 | 468.4 | – | – | 65536 | 2026-10-07 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.5 | 208.4 | – | – | 1024 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.5 | 208.4 | – | – | 1024 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.5 | 208.4 | – | – | 1024 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.5 | 208.4 | – | – | 1024 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.5 | 208.4 | – | – | 1024 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.5 | 208.4 | – | – | 1024 | 2026-10-11 | community | contradiction | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 20.3 | 85.2 | – | – | 1024 | 2026-10-09 | community | – | source |
| M4 Pro | gemma-4-31B-it | oQ4e | oMLX | 20.2 | 104.1 | – | – | 4096 | 2026-10-17 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.1 | 158.0 | – | – | 65536 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.1 | 158.0 | – | – | 65536 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.1 | 158.0 | – | – | 65536 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.1 | 158.0 | – | – | 65536 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.1 | 158.0 | – | – | 65536 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 20.1 | 158.0 | – | – | 65536 | 2026-10-11 | community | – | source |
| M4 Pro | gemma-4-31B-it | oQ4e | oMLX | 19.9 | 105.5 | – | – | 1024 | 2026-10-17 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 19.8 | 178.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 19.8 | 178.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 19.8 | 178.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 19.8 | 178.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 19.8 | 178.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 19.8 | 178.0 | – | – | 32768 | 2026-10-11 | community | contradiction | source |
| M2 Ultra | gemma-4-31B-it | 4bit | oMLX | 19.7 | 191.2 | – | – | 1024 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 19.4 | 533.7 | – | – | 65536 | 2026-10-12 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 19.1 | 180.0 | – | – | 1024 | 2026-10-12 | community | contradiction | source |
| M2 Ultra | gemma-4-31B-it | 4bit | oMLX | 19.0 | 185.7 | – | – | 4096 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | 6bit | oMLX | 18.8 | 644.7 | – | – | 4096 | 2026-10-13 | community | – | source |
| M3 Ultra | gemma-4-31B-it | 8bit | oMLX | 18.5 | 341.5 | – | – | 1024 | 2026-10-11 | community | – | source |
| M3 Ultra | gemma-4-31B-it | 8bit | oMLX | 18.5 | 341.5 | – | – | 1024 | 2026-10-11 | community | – | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 18.5 | 120.7 | – | – | 4096 | 2026-10-12 | community | contradiction | source |
| M1 Max | gemma-4-31B-it | oQ4e | oMLX | 18.5 | 89.3 | – | – | 4096 | 2026-10-18 | community | – | source |
| M3 Ultra | gemma-4-31B-it | 8bit | oMLX | 18.3 | 339.6 | – | – | 4096 | 2026-10-11 | community | – | source |
| M3 Ultra | gemma-4-31B-it | 8bit | oMLX | 18.3 | 339.6 | – | – | 4096 | 2026-10-11 | community | – | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 18.0 | 84.1 | – | – | 4096 | 2026-10-09 | community | – | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 17.9 | 83.8 | – | – | 4096 | 2026-10-13 | community | – | source |
| M3 Ultra | gemma-4-31B-it | 8bit | oMLX | 17.8 | 336.2 | – | – | 8192 | 2026-10-11 | community | – | source |
| M3 Ultra | gemma-4-31B-it | 8bit | oMLX | 17.8 | 336.2 | – | – | 8192 | 2026-10-11 | community | – | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 17.6 | 85.3 | – | – | 1024 | 2026-10-09 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 6bit | oMLX | 17.4 | 510.4 | – | – | 16384 | 2026-10-13 | community | – | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 155.1 | – | – | 65536 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 155.1 | – | – | 65536 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 155.1 | – | – | 65536 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 155.1 | – | – | 65536 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 155.1 | – | – | 65536 | 2026-10-11 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 155.1 | – | – | 65536 | 2026-10-11 | community | contradiction | source |
| M2 Max | gemma-4-31B-it | oQ4e | oMLX | 17.0 | 91.9 | – | – | 4096 | 2026-10-17 | community | – | source |
| M2 Max | gemma-4-31B-it | oQ4e | oMLX | 15.9 | 102.8 | – | – | 1024 | 2026-10-17 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 15.8 | 503.1 | – | – | 65536 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ6e | oMLX | 15.8 | 503.1 | – | – | 65536 | 2026-10-12 | community | contradiction | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 15.5 | 85.2 | – | – | 1024 | 2026-10-13 | community | contradiction | source |
| M2 Max | gemma-4-31B-it | oQ4e | oMLX | 15.5 | 89.0 | – | – | 1024 | 2026-10-18 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 15.4 | 645.0 | – | – | 1024 | 2026-10-17 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 15.4 | 645.0 | – | – | 1024 | 2026-10-17 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 15.3 | 640.4 | – | – | 1024 | 2026-10-16 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 15.1 | 411.8 | – | – | 131072 | 2026-10-12 | community | – | source |
| M5 Max | gemma-4-31B-it | oQ4e | oMLX | 15.0 | 327.5 | – | – | 131072 | 2026-10-07 | community | – | source |
| M1 Ultra | gemma-4-31B-it | mxfp4 | oMLX | 15.0 | 167.0 | – | – | 16384 | 2026-10-09 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 15.0 | 699.9 | – | – | 4096 | 2026-10-17 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 15.0 | 699.9 | – | – | 4096 | 2026-10-17 | community | contradiction | source |
| M5 Pro | gemma-4-31B-it | 4bit | oMLX | 14.7 | 377.4 | – | – | 1024 | 2026-10-13 | community | – | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 14.6 | 81.2 | – | – | 8192 | 2026-10-13 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 14.5 | 687.4 | – | – | 8192 | 2026-10-17 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 14.5 | 687.4 | – | – | 8192 | 2026-10-17 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 14.4 | 637.4 | – | – | 1024 | 2026-10-18 | community | contradiction | source |
| M1 Ultra | gemma-4-31B-it | oQ6e | oMLX | 14.3 | 166.8 | – | – | 16384 | 2026-10-09 | community | – | source |
| M5 Pro | gemma-4-31B-it | 4bit | oMLX | 14.1 | 342.3 | – | – | 4096 | 2026-10-13 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 14.1 | 531.2 | – | – | 4096 | 2026-10-18 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 13.6 | 662.8 | – | – | 16384 | 2026-10-17 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 13.6 | 662.8 | – | – | 16384 | 2026-10-17 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 13.6 | 471.0 | – | – | 8192 | 2026-10-18 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 13.5 | 496.4 | – | – | 16384 | 2026-10-09 | community | contradiction | source |
| M4 Pro | gemma-4-31B-it | oQ8e | oMLX | 13.2 | 83.7 | – | – | 4096 | 2026-10-09 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 13.1 | 435.8 | – | – | 16384 | 2026-10-18 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 13.0 | 160.6 | – | – | 4096 | 2026-10-12 | community | contradiction | source |
| M5 Pro | gemma-4-31B-it | oQ4e | oMLX | 12.9 | 369.9 | – | – | 8192 | 2026-10-07 | community | – | source |
| M4 Pro | gemma-4-31B-it | 4bit | oMLX | 12.8 | 107.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| M4 Pro | gemma-4-31B-it | 4bit | oMLX | 12.6 | 103.2 | – | – | 4096 | 2026-10-15 | community | – | source |
| M4 Max | gemma-4-31B-it | 4bit | oMLX | 12.5 | 186.2 | – | – | 4096 | 2026-10-12 | community | contradiction | source |
| M4 Max | gemma-4-31B-it | 8bit | oMLX | 12.5 | 140.8 | – | – | 1024 | 2026-10-12 | community | contradiction | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 12.4 | 456.1 | – | – | 65536 | 2026-10-16 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 10.3 | 340.8 | – | – | 131072 | 2026-10-16 | community | – | source |
| M2 Max | gemma-4-31B-it | 8bit | oMLX | 10.2 | 116.5 | – | – | 1024 | 2026-10-08 | community | – | source |
| M2 Max | gemma-4-31B-it | 8bit | oMLX | 10.1 | 111.9 | – | – | 4096 | 2026-10-08 | community | – | source |
| M5 Pro | gemma-4-31B-it | oQ4e | oMLX | 9.7 | 289.4 | – | – | 32768 | 2026-10-07 | community | – | source |
| M5 Max | gemma-4-31B-it | 8bit | oMLX | 8.2 | 266.3 | – | – | 199680 | 2026-10-16 | community | – | source |
| M1 Max | gemma-4-31B-it | 8bit | oMLX | 7.9 | 72.2 | – | – | 1024 | 2026-10-11 | community | – | source |
| M1 Max | gemma-4-31B-it | 8bit | oMLX | 7.9 | 72.2 | – | – | 1024 | 2026-10-11 | community | – | source |
| M1 Max | gemma-4-31B-it | 8bit | oMLX | 7.3 | 67.1 | – | – | 4096 | 2026-10-11 | community | – | source |
| M1 Max | gemma-4-31B-it | 8bit | oMLX | 7.3 | 67.1 | – | – | 4096 | 2026-10-11 | community | – | source |
| M5 | gemma-4-31B-it | mxfp4 | oMLX | 6.9 | 169.0 | – | – | 1024 | 2026-10-15 | community | – | source |
| AMD Custom GPU 0932 | gemma-4-31B-it 30.70B | IQ2_XXS - 2.0625 bpw | llama.cpp (Vulkan) | 3.56 | 17.22 | – | – | tg128 | 2026-08-20 | community | – | source |
| AMD Custom GPU 0932 | gemma-4-31B-it 30.70B | IQ2_XXS - 2.0625 bpw | llama.cpp (Vulkan) | 3.50 | 17.15 | – | – | tg128 | 2026-08-20 | community | – | source |
| RTX 5060 Ti | gemma-4-31B-it 31B | Q6_K | llama.cpp (CUDA) | 0.62 | 41.84 | – | – | – | 2026-08 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 3566.7 | – | – | 1024 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 3835.8 | – | – | 1024 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 1906.1 | – | – | 512 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 3173.0 | – | – | 512 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 398.2 | – | – | 128 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 1266.7 | – | – | 128 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 3306.2 | – | – | 4096 | 2026-10 | community | – | source |
| A100-SXM4-80GB | gemma-4-31B-it | BF16 | vLLM | – | 3390.3 | – | – | 4096 | 2026-10 | community | – | source |