4x H200 NVL
11 records across 1 models (quants: NVFP4).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | 823.9 | – | – | – | – | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | 761.1 | – | – | – | – | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | 329.8 | – | – | – | – | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | 279.2 | – | – | – | – | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | 119.1 | – | – | – | – | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | 110.7 | – | – | – | – | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | – | 5092 | – | – | 32768 | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | – | 5129 | – | – | 32768 | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | – | 4941 | – | – | 131072 | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | – | 4988 | – | – | 131072 | 2026-09-30 | community | – | source |
| 4x H200 NVL | GLM-5.3 | NVFP4 | vllm | – | 5103 | – | – | 32768 | 2026-09-30 | community | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085