hw_id stringclasses 12
values | device_raw stringclasses 12
values | device_count int64 1 1 | backend stringclasses 1
value | model stringclasses 13
values | quant stringclasses 2
values | params_b float64 4.66 117 | size_gib float64 2.54 68 | fa int64 1 1 | ngl int64 99 99 | threads int64 6 128 | ubatch int64 512 512 | depth int64 0 131k | pp_n int64 512 512 | pp_tps float64 290 20.5k | tg_n int64 128 128 | tg_tps float64 4.18 396 | build stringclasses 1
value | build_num int64 11.7k 11.7k | n_samples int64 3 3 | date stringdate 2026-10-16 00:00:00 2026-10-21 00:00:00 | notes stringlengths 295 507 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 12 | 512 | 0 | 512 | 1,277.12 | 128 | 39.71 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 17.55; tg ... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 12 | 512 | 4,096 | 512 | 1,151.05 | 128 | 37.73 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 11.18; tg ... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 12 | 512 | 16,384 | 512 | 950.78 | 128 | 36.79 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 6.81; tg s... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 12 | 512 | 0 | 512 | 598.35 | 128 | 41.12 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp s... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 12 | 512 | 0 | 512 | 476.43 | 128 | 34.64 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp s... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 12 | 512 | 4,096 | 512 | 583.06 | 128 | 39.55 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp s... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 12 | 512 | 16,384 | 512 | 557.1 | 128 | 38.67 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp s... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 12 | 512 | 0 | 512 | 476.61 | 128 | 45.66 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instruc... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 12 | 512 | 0 | 512 | 365.02 | 128 | 35.31 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instruc... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 12 | 512 | 4,096 | 512 | 471.52 | 128 | 42.54 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instruc... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 12 | 512 | 16,384 | 512 | 400.83 | 128 | 37.82 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instruc... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 12 | 512 | 0 | 512 | 1,983.02 | 128 | 61.52 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 12 | 512 | 4,096 | 512 | 1,708.77 | 128 | 55.39 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 12 | 512 | 16,384 | 512 | 1,182.57 | 128 | 43.02 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-VL 8B | Q8_0 | 8.77 | 8.11 | 1 | 99 | 12 | 512 | 0 | 512 | 2,086.63 | 128 | 39.25 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3vl 8B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q8_0.gguf; pp stddev 42.87; t... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-VL 8B | Q8_0 | 8.77 | 8.11 | 1 | 99 | 12 | 512 | 4,096 | 512 | 1,776.4 | 128 | 36.59 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3vl 8B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q8_0.gguf; pp stddev 15.71; t... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3-VL 8B | Q8_0 | 8.77 | 8.11 | 1 | 99 | 12 | 512 | 16,384 | 512 | 1,215.46 | 128 | 30.64 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen3vl 8B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q8_0.gguf; pp stddev 7.60; tg... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 12 | 512 | 0 | 512 | 2,996.74 | 128 | 85.27 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 115.70; tg stddev ... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 12 | 512 | 4,096 | 512 | 2,843.96 | 128 | 83.15 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 46.46; tg stddev 0... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 12 | 512 | 16,384 | 512 | 2,425.84 | 128 | 75.86 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 29.74; tg stddev 0... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 4B | Q8_0 | 4.66 | 4.16 | 1 | 99 | 12 | 512 | 0 | 512 | 3,133.42 | 128 | 62.6 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 4B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q8_0.gguf; pp stddev 107.71; tg stddev 0.12 |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 4B | Q8_0 | 4.66 | 4.16 | 1 | 99 | 12 | 512 | 4,096 | 512 | 2,992.45 | 128 | 61.27 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 4B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q8_0.gguf; pp stddev 49.17; tg stddev 0.10 |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 4B | Q8_0 | 4.66 | 4.16 | 1 | 99 | 12 | 512 | 16,384 | 512 | 2,506.95 | 128 | 57.05 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 4B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q8_0.gguf; pp stddev 40.31; tg stddev 0.08 |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 12 | 512 | 0 | 512 | 1,849.31 | 128 | 55.28 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 43.17; tg stddev 0... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 12 | 512 | 4,096 | 512 | 1,800.39 | 128 | 54.23 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 18.40; tg stddev 0... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 12 | 512 | 16,384 | 512 | 1,613.1 | 128 | 50.98 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 14.50; tg stddev 0... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 9B | Q8_0 | 9.65 | 8.86 | 1 | 99 | 12 | 512 | 0 | 512 | 1,945.31 | 128 | 36.6 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 9B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q8_0.gguf; pp stddev 47.53; tg stddev 0.04 |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 9B | Q8_0 | 9.65 | 8.86 | 1 | 99 | 12 | 512 | 4,096 | 512 | 1,891.49 | 128 | 36.12 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 9B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q8_0.gguf; pp stddev 18.52; tg stddev 0.04 |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.5 9B | Q8_0 | 9.65 | 8.86 | 1 | 99 | 12 | 512 | 16,384 | 512 | 1,683.59 | 128 | 34.62 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35 9B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q8_0.gguf; pp stddev 16.59; tg stddev 0.03 |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.6 35B-A3B | Q4_K_M | 35.95 | 20.6 | 1 | 99 | 12 | 512 | 0 | 512 | 360.8 | 128 | 41.13 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35moe 35B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-35B-A3B-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf; pp stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.6 35B-A3B | Q4_K_M | 35.95 | 20.6 | 1 | 99 | 12 | 512 | 4,096 | 512 | 342.37 | 128 | 40.74 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35moe 35B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-35B-A3B-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf; pp stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | Qwen3.6 35B-A3B | Q4_K_M | 35.95 | 20.6 | 1 | 99 | 12 | 512 | 16,384 | 512 | 328.8 | 128 | 39.76 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'qwen35moe 35B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-35B-A3B-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf; pp stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 12 | 512 | 0 | 512 | 1,077.18 | 128 | 57.61 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 34.32; tg stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 12 | 512 | 0 | 512 | 785.07 | 128 | 43.2 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 30.38; tg stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 12 | 512 | 0 | 512 | 616.4 | 128 | 32.52 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 21.36; tg stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 12 | 512 | 4,096 | 512 | 1,067.79 | 128 | 56.01 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 14.82; tg stdd... |
rtx-3060-12 | NVIDIA GeForce RTX 3060 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 12 | 512 | 16,384 | 512 | 969.92 | 128 | 53.71 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.91.07, vast.ai, CPU AMD Ryzen Threadripper 1920X 12-Core Processor, RAM 62 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 4.17; tg stdde... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 24 | 512 | 0 | 512 | 3,271.97 | 128 | 86.54 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 87.38; t... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 24 | 512 | 4,096 | 512 | 2,960.69 | 128 | 82.7 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 49.26; t... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 24 | 512 | 16,384 | 512 | 2,438.93 | 128 | 80.19 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 30.26; t... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q8_0 | 11.96 | 11.78 | 1 | 99 | 24 | 512 | 0 | 512 | 3,408.05 | 128 | 57.24 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 ?B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q8_0.gguf; pp stddev 80.86; tg stddev 0.... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q8_0 | 11.96 | 11.78 | 1 | 99 | 24 | 512 | 4,096 | 512 | 3,080.06 | 128 | 55.48 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 ?B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q8_0.gguf; pp stddev 50.13; tg stddev 0.... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q8_0 | 11.96 | 11.78 | 1 | 99 | 24 | 512 | 16,384 | 512 | 2,523.11 | 128 | 54.32 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 ?B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q8_0.gguf; pp stddev 42.38; tg stddev 0.... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 24 | 512 | 0 | 512 | 4,663.86 | 128 | 140.26 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 24 | 512 | 4,096 | 512 | 4,285.15 | 128 | 132.74 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 26B-A4B | Q4_K_M | 25.81 | 15.77 | 1 | 99 | 24 | 512 | 16,384 | 512 | 3,667.99 | 128 | 127.42 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 26B.A4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-26B-A4B-it-GGUF/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf; pp... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 31B | Q4_K_M | 31.27 | 17.05 | 1 | 99 | 24 | 512 | 0 | 512 | 1,292.43 | 128 | 38.78 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 31B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-31B-it-GGUF/gemma-4-31B-it-Q4_K_M.gguf; pp stddev 11.98; ... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 31B | Q4_K_M | 31.27 | 17.05 | 1 | 99 | 24 | 512 | 4,096 | 512 | 1,182.19 | 128 | 36.8 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 31B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-31B-it-GGUF/gemma-4-31B-it-Q4_K_M.gguf; pp stddev 8.98; t... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 31B | Q4_K_M | 31.27 | 17.05 | 1 | 99 | 24 | 512 | 16,384 | 512 | 968.76 | 128 | 35.23 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gemma4 31B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-31B-it-GGUF/gemma-4-31B-it-Q4_K_M.gguf; pp stddev 6.11; t... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Mistral Small 3.2 24B | Q4_K_M | 24.01 | 13.34 | 1 | 99 | 24 | 512 | 0 | 512 | 1,897.06 | 128 | 53.15 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'llama 13B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/Mistral-Small-3.2-24B-Ins... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Mistral Small 3.2 24B | Q4_K_M | 24.01 | 13.34 | 1 | 99 | 24 | 512 | 4,096 | 512 | 1,793.83 | 128 | 51.07 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'llama 13B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/Mistral-Small-3.2-24B-Ins... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Mistral Small 3.2 24B | Q4_K_M | 24.01 | 13.34 | 1 | 99 | 24 | 512 | 16,384 | 512 | 1,473.55 | 128 | 45.91 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'llama 13B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/Mistral-Small-3.2-24B-Ins... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 24 | 512 | 0 | 512 | 4,304.46 | 128 | 206.29 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instr... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 24 | 512 | 4,096 | 512 | 3,709.85 | 128 | 188.55 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instr... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 24 | 512 | 16,384 | 512 | 2,381.21 | 128 | 149.37 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instr... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 24 | 512 | 0 | 512 | 4,898.11 | 128 | 134.6 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 24 | 512 | 4,096 | 512 | 4,285.15 | 128 | 123.07 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 24 | 512 | 16,384 | 512 | 2,940.27 | 128 | 98.69 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q8_0 | 8.77 | 8.11 | 1 | 99 | 24 | 512 | 0 | 512 | 5,109.2 | 128 | 91.35 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3vl 8B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q8_0.gguf; pp stddev 134.56... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q8_0 | 8.77 | 8.11 | 1 | 99 | 24 | 512 | 4,096 | 512 | 4,418.49 | 128 | 85.86 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3vl 8B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q8_0.gguf; pp stddev 95.27;... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q8_0 | 8.77 | 8.11 | 1 | 99 | 24 | 512 | 16,384 | 512 | 2,997.95 | 128 | 73.05 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen3vl 8B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q8_0.gguf; pp stddev 43.61;... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 24 | 512 | 0 | 512 | 7,365.78 | 128 | 179.09 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 449.72; tg stdde... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 24 | 512 | 4,096 | 512 | 7,162.07 | 128 | 175.92 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 250.20; tg stdde... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 24 | 512 | 16,384 | 512 | 5,994.92 | 128 | 163.16 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 147.61; tg stdde... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q8_0 | 4.66 | 4.16 | 1 | 99 | 24 | 512 | 0 | 512 | 7,630.74 | 128 | 137.98 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 4B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q8_0.gguf; pp stddev 499.34; tg stddev 0.15 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q8_0 | 4.66 | 4.16 | 1 | 99 | 24 | 512 | 4,096 | 512 | 7,371.12 | 128 | 135.64 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 4B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q8_0.gguf; pp stddev 232.40; tg stddev 0.27 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q8_0 | 4.66 | 4.16 | 1 | 99 | 24 | 512 | 16,384 | 512 | 6,231.5 | 128 | 127.55 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 4B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q8_0.gguf; pp stddev 184.47; tg stddev 0.34 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 24 | 512 | 0 | 512 | 4,632.1 | 128 | 122.08 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 160.82; tg stdde... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 24 | 512 | 4,096 | 512 | 4,544.73 | 128 | 120.15 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 83.25; tg stddev... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 24 | 512 | 16,384 | 512 | 4,027.01 | 128 | 113.85 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 65.33; tg stddev... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q8_0 | 9.65 | 8.86 | 1 | 99 | 24 | 512 | 0 | 512 | 4,788.32 | 128 | 85.66 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 9B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q8_0.gguf; pp stddev 169.58; tg stddev 0.10 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q8_0 | 9.65 | 8.86 | 1 | 99 | 24 | 512 | 4,096 | 512 | 4,718.64 | 128 | 84.76 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 9B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q8_0.gguf; pp stddev 82.15; tg stddev 0.14 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q8_0 | 9.65 | 8.86 | 1 | 99 | 24 | 512 | 16,384 | 512 | 4,216.81 | 128 | 81.48 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 9B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q8_0.gguf; pp stddev 80.41; tg stddev 0.12 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 27B | Q4_K_M | 27.78 | 15.65 | 1 | 99 | 24 | 512 | 0 | 512 | 1,515.77 | 128 | 42.18 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 27B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-27B-GGUF/Qwen3.6-27B-Q4_K_M.gguf; pp stddev 19.65; tg std... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 27B | Q4_K_M | 27.78 | 15.65 | 1 | 99 | 24 | 512 | 4,096 | 512 | 1,462.74 | 128 | 41.69 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 27B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-27B-GGUF/Qwen3.6-27B-Q4_K_M.gguf; pp stddev 15.01; tg std... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 27B | Q4_K_M | 27.78 | 15.65 | 1 | 99 | 24 | 512 | 16,384 | 512 | 1,276.35 | 128 | 40.1 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35 27B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-27B-GGUF/Qwen3.6-27B-Q4_K_M.gguf; pp stddev 16.82; tg std... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 35B-A3B | Q4_K_M | 35.95 | 20.6 | 1 | 99 | 24 | 512 | 0 | 512 | 4,053.33 | 128 | 158.97 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35moe 35B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-35B-A3B-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf; pp st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 35B-A3B | Q4_K_M | 35.95 | 20.6 | 1 | 99 | 24 | 512 | 4,096 | 512 | 3,868.21 | 128 | 157.6 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35moe 35B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-35B-A3B-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf; pp st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 35B-A3B | Q4_K_M | 35.95 | 20.6 | 1 | 99 | 24 | 512 | 16,384 | 512 | 3,430 | 128 | 150.59 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'qwen35moe 35B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-35B-A3B-GGUF/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf; pp st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 24 | 512 | 0 | 512 | 5,857.32 | 128 | 214.34 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 282.22; tg s... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 24 | 512 | 4,096 | 512 | 5,647.65 | 128 | 208.37 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 94.54; tg st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | gpt-oss 20B | Q4_K_M | 20.91 | 10.81 | 1 | 99 | 24 | 512 | 16,384 | 512 | 4,546.31 | 128 | 193.84 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gpt-oss 20B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q4_K_M.gguf; pp stddev 95.03; tg st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | gpt-oss 20B | Q8_0 | 20.91 | 11.27 | 1 | 99 | 24 | 512 | 0 | 512 | 5,951.86 | 128 | 204.22 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gpt-oss 20B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q8_0.gguf; pp stddev 252.23; tg stddev 1.07 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | gpt-oss 20B | Q8_0 | 20.91 | 11.27 | 1 | 99 | 24 | 512 | 4,096 | 512 | 5,696.08 | 128 | 198.28 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gpt-oss 20B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q8_0.gguf; pp stddev 97.20; tg stddev 1.15 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | gpt-oss 20B | Q8_0 | 20.91 | 11.27 | 1 | 99 | 24 | 512 | 16,384 | 512 | 4,521.59 | 128 | 185.17 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-16 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 580.126.09, vast.ai, CPU AMD Ryzen Threadripper 3960X 24-Core Processor, RAM 125 GiB; llama-bench model_type 'gpt-oss 20B Q8_0'; -r 3; GPU 0 only; GGUF unsloth/gpt-oss-20b-GGUF/gpt-oss-20b-Q8_0.gguf; pp stddev 96.17; tg stddev 0.73 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 96 | 512 | 32,768 | 512 | 1,931.21 | 128 | 78.39 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 57.76; tg stddev 0.26 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 96 | 512 | 65,536 | 512 | 1,407.72 | 128 | 74.7 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 25.32; tg stddev 0.20 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Gemma 4 12B | Q4_K_M | 11.96 | 6.62 | 1 | 99 | 96 | 512 | 131,072 | 512 | 904.93 | 128 | 67.76 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'gemma4 ?B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/gemma-4-12b-it-GGUF/gemma-4-12b-it-Q4_K_M.gguf; pp stddev 10.52; tg stddev 0.16 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Mistral Small 3.2 24B | Q4_K_M | 24.01 | 13.34 | 1 | 99 | 96 | 512 | 32,768 | 512 | 1,188.33 | 128 | 40.65 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'llama 13B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/Mistral-Small-3.2-24B-Instruct-2506-Q4_K_... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-Coder 30B-A3B | Q4_K_M | 30.53 | 17.28 | 1 | 99 | 96 | 512 | 32,768 | 512 | 1,609.47 | 128 | 116.4 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen3moe 30B.A3B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF/Qwen3-Coder-30B-A3B-Instruct-Q4_K_M.gguf;... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 96 | 512 | 32,768 | 512 | 2,014.44 | 128 | 77.68 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp stddev 45.34; tg s... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 96 | 512 | 65,536 | 512 | 1,284.16 | 128 | 54.47 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp stddev 13.11; tg s... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3-VL 8B | Q4_K_M | 8.77 | 4.68 | 1 | 99 | 96 | 512 | 131,072 | 512 | 739.68 | 128 | 34.06 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen3vl 8B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3-VL-8B-Instruct-GGUF/Qwen3-VL-8B-Instruct-Q4_K_M.gguf; pp stddev 6.60; tg st... |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 96 | 512 | 32,768 | 512 | 4,576.75 | 128 | 146.97 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 257.83; tg stddev 1.02 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 96 | 512 | 65,536 | 512 | 3,454.06 | 128 | 123.77 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 118.07; tg stddev 0.39 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 4B | Q4_K_M | 4.66 | 2.54 | 1 | 99 | 96 | 512 | 131,072 | 512 | 2,254.38 | 128 | 93.97 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 4B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-4B-GGUF/Qwen3.5-4B-Q4_K_M.gguf; pp stddev 85.81; tg stddev 0.27 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 96 | 512 | 32,768 | 512 | 3,413.32 | 128 | 106.21 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 146.13; tg stddev 0.25 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 96 | 512 | 65,536 | 512 | 2,740.12 | 128 | 93.79 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 51.28; tg stddev 0.25 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.5 9B | Q4_K_M | 9.65 | 5.28 | 1 | 99 | 96 | 512 | 131,072 | 512 | 1,903.9 | 128 | 75.84 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 9B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.5-9B-GGUF/Qwen3.5-9B-Q4_K_M.gguf; pp stddev 47.53; tg stddev 0.23 |
rtx-3090 | NVIDIA GeForce RTX 3090 | 1 | CUDA | Qwen3.6 27B | Q4_K_M | 27.78 | 15.65 | 1 | 99 | 96 | 512 | 32,768 | 512 | 1,081.03 | 128 | 38.48 | 3e4dfa8f3 | 11,658 | 3 | 2026-10-18 | own run, llama-bench from llama.cpp b11658 (commit 3e4dfa8f3), CUDA 12.8 build, driver 595.71.05, vast.ai, CPU AMD EPYC 7642 48-Core Processor, RAM 251 GiB; llama-bench model_type 'qwen35 27B Q4_K - Medium'; -r 3; GPU 0 only; GGUF unsloth/Qwen3.6-27B-GGUF/Qwen3.6-27B-Q4_K_M.gguf; pp stddev 26.66; tg stddev 0.06 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.