🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the NVIDIA B200 run GLM-4V-9B?
Yes — here's how fast and up to what context, with real measured numbers.
GLM-4V-9B on the NVIDIA B200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1199k | 324.4-386.7 t/s | ~1.6-1.9 (slow) |
| Q8 | yes | 1268k | 648.9-773.3 t/s | ~3.2-3.9 (slow) |
| Q4 | yes | 1298k | 1158.7-1381 t/s | ~5.8-6.9 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to system RAM offload (depends on your system RAM) and speed drops to the "beyond" figure.
FAQ
Can the NVIDIA B200 run GLM-4V-9B?▶
Yes. At Q4 it generates ~1158.7-1381 tok/s and fits up to 1298k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~5.8-6.9 tok/s.
How much context fits?▶
At Q4, up to about 1298k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.