Yes — with 59.5 GB to spare
Granite 4.1 30B at Q4_K_M fits your H100 PCIe entirely on the GPU at 8K context, at an estimated 75 tokens per second. There is room for its full 128K window.
Fully on GPU
8K context
Q4_K_M · 16.3 GB
Apache 2.0
Released 29 Apr 2026
The largest Granite. Dense 29B at Q4 is a comfortable 24 GB fit.
The VRAM budget
weights 16.3 GB
Weights 16.3 GB
KV cache @ 8K 2.00 GB
Runtime overhead 0.6 GB
Free 59.5 GB of 78.4 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| Q8_0 | 28.6 GB | 31.2 GB | 128K | 42 | −0.1% ppl | Long context |
| Q6_K | 22.1 GB | 24.7 GB | 128K | 55 | −0.4% ppl | Long context |
| Q5_K_M | 19.1 GB | 21.7 GB | 128K | 63 | −0.8% ppl | Long context |
| Q4_K_M | 16.3 GB | 18.9 GB | 128K | 75 | −1.9% ppl | Recommended |
| Q3_K_M | 13.2 GB | 15.8 GB | 128K | 92 | −5.4% ppl | Long context |
| Q2_K | 11.3 GB | 13.9 GB | 128K | 107 | −15% ppl | Long context |
Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it.
How to run it
$ ollama pull granite4.1:30b $ OLLAMA_CONTEXT_LENGTH=8192 \ ollama run granite4.1:30b
The default. One binary, a model registry, an OpenAI-compatible port. More on Ollama.
01Download is 16.3 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.
02Close anything else holding VRAM. A browser with hardware acceleration can sit on 1–2 GB.
03There is room to go to the model's full 128K context on this card.