🔌
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 H200 run Pleias-RAG-1B?
Yes — here's how fast and up to what context, with real measured numbers.
Pleias-RAG-1B on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1920k | 1460-1740 t/s | ~12.2-14.5 (slow) |
| Q8 | yes | 1938k | 2920-3480 t/s | ~24.3-29 (slow) |
| Q4 | yes | 1946k | 5214.3-6214.3 t/s | ~43.5-51.8 (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 H200 run Pleias-RAG-1B?▶
Yes. At Q4 it generates ~5214.3-6214.3 tok/s and fits up to 1946k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~43.5-51.8 tok/s.
How much context fits?▶
At Q4, up to about 1946k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.