| model                          |       size |     params | backend    | ngl | type_k | type_v |  fa |            test |                  t/s |
| ------------------------------ | ---------: | ---------: | ---------- | --: | -----: | -----: | --: | --------------: | -------------------: |
| qwen3next 80B.A3B Q8_0         |  78.98 GiB |    79.67 B | CUDA       | 999 |   q8_0 |   q8_0 |   1 |           pp512 |      4865.41 ± 16.43 |
| qwen3next 80B.A3B Q8_0         |  78.98 GiB |    79.67 B | CUDA       | 999 |   q8_0 |   q8_0 |   1 |          pp4096 |       4889.55 ± 6.40 |
| qwen3next 80B.A3B Q8_0         |  78.98 GiB |    79.67 B | CUDA       | 999 |   q8_0 |   q8_0 |   1 |           tg128 |        195.11 ± 0.86 |
| qwen3next 80B.A3B Q8_0         |  78.98 GiB |    79.67 B | CUDA       | 999 |   q8_0 |   q8_0 |   1 |  pp512 @ d16384 |      4627.08 ± 24.10 |
| qwen3next 80B.A3B Q8_0         |  78.98 GiB |    79.67 B | CUDA       | 999 |   q8_0 |   q8_0 |   1 | pp4096 @ d16384 |       4611.38 ± 5.30 |
| qwen3next 80B.A3B Q8_0         |  78.98 GiB |    79.67 B | CUDA       | 999 |   q8_0 |   q8_0 |   1 |  tg128 @ d16384 |        183.73 ± 1.23 |

build: d2462f8f7 (9590)
