GeForce RTX 3090
46 records across 3 models (quants: Q4_0, Q5_K - Medium).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 177.79 | 5992.1 | – | – | tg128 | 2026-05-03 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 177.75 | 5973.72 | – | – | tg128 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 175.48 | 5874.09 | – | – | tg128 | 2025-12-11 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 168.95 | 5153.91 | – | – | tg128 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 168.83 | 5349.43 | – | – | tg128 | 2026-05-03 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 167.39 | 5134.61 | – | – | tg128 | 2025-12-11 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 160.67 | 5051.48 | – | – | tg128 | 2025-12-11 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 155.36 | 5932.44 | – | – | tg128 | 2025-08-02 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 153.01 | 4425.82 | – | – | tg128 | 2025-12-11 | community | – | source |
| GeForce RTX 3090 | llama 7B | Q4_0 | llama.cpp (CUDA) | 151.04 | 5405.83 | – | – | tg128 | 2025-08-02 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 77.24 | 1715.55 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 75.34 | 2045.61 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 75.18 | 2045.01 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 74.75 | 2047.57 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 74.61 | 2050.92 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 74.48 | 2051.13 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 39.0 | 1403.46 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.99 | 1401.5 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.99 | 1401.56 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.98 | 1397.03 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.98 | 1404.01 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.98 | 1403.68 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.57 | 1379.75 | – | – | tg512 | 2026-09-07 | community | contradiction | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.57 | 1377.26 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.56 | 1382.0 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.54 | 1380.63 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.54 | 1379.84 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35 27B | Q5_K - Medium | llama.cpp (CUDA) | 38.54 | 1382.78 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 158.25 | 4384.76 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 157.32 | 4305.41 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 157.14 | 4424.02 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 156.64 | 4358.85 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 156.16 | 4058.5 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 151.13 | 3600.54 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 151.01 | 3601.63 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 150.93 | 3583.97 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 150.48 | 3580.28 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 150.4 | 3579.26 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 150.31 | 3619.07 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 148.79 | 3530.83 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 148.69 | 3549.89 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 148.5 | 3560.95 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 148.41 | 3586.48 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 148.36 | 3587.97 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 148.13 | 3531.31 | – | – | tg512 | 2026-09-07 | community | – | source |
| GeForce RTX 3090 | qwen35moe 35B.A3B | Q5_K - Medium | llama.cpp (CUDA) | 147.76 | 4364.06 | – | – | tg512 | 2026-09-07 | community | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085