Falcon-H1-7B
10 records across 2 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| RTX PRO 6000 | Falcon-H1-7B 7B | Q4_K_M | llama.cpp (CUDA) | 1417.1 | – | – | – | – | 2026-08 | community | – | source |
| RTX PRO 6000 | Falcon-H1-7B 7B | Q8_0 | llama.cpp (CUDA) | 1318.3 | – | – | – | – | 2026-08 | community | – | source |
| RTX PRO 6000 | Falcon-H1-7B 7B | BF16 | llama.cpp (CUDA) | 1188.2 | – | – | – | – | 2026-08 | community | – | source |
| RTX PRO 6000 | Falcon-H1-7B 7B | Q4_K_M | llama.cpp (CUDA) | 1067.9 | – | – | – | – | 2026-08 | community | – | source |
| RTX PRO 6000 | Falcon-H1-7B 7B | Q8_0 | llama.cpp (CUDA) | 963.3 | – | – | – | – | 2026-08 | community | – | source |
| RTX PRO 6000 | Falcon-H1-7B 7B | BF16 | llama.cpp (CUDA) | 771.5 | – | – | – | – | 2026-08 | community | – | source |
| GB10 | Falcon-H1-7B 7B | Q4_K_M | llama.cpp (CUDA) | 273.1 | – | – | – | – | 2026-08 | community | – | source |
| GB10 | Falcon-H1-7B 7B | BF16 | llama.cpp (CUDA) | 199.0 | – | – | – | – | 2026-08 | community | – | source |
| GB10 | Falcon-H1-7B 7B | Q4_K_M | llama.cpp (CUDA) | 187.7 | – | – | – | – | 2026-08 | community | – | source |
| GB10 | Falcon-H1-7B 7B | BF16 | llama.cpp (CUDA) | 80.0 | – | – | – | – | 2026-08 | community | – | source |