ibm-granite/granite-4.1-8b
ibm-granite/granite-4.1-8b
Parameters
8.5B
Context
128K
Downloads
4.1M
Likes
249
Architecture
granite
40 layers · 8 KV heads · head 128 · hidden 4096
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 25.6 GB RAM 38.4 · Disk 63.6 | 46.2 GB RAM 69.2 · Disk 89.6 | 168.2 GB RAM 252.2 · Disk 128.6 |
| FP88bit | 15.6 GB RAM 32.0 · Disk 51.3 | 35.7 GB RAM 53.6 · Disk 77.3 | 157.3 GB RAM 235.9 · Disk 116.3 |
| INT88bit | 15.6 GB RAM 32.0 · Disk 51.3 | 35.7 GB RAM 53.6 · Disk 77.3 | 157.3 GB RAM 235.9 · Disk 116.3 |
| AWQ4bit | 10.8 GB RAM 32.0 · Disk 45.4 | 30.7 GB RAM 46.1 · Disk 71.4 | 152.0 GB RAM 228.1 · Disk 110.4 |
| GPTQ4bit | 10.9 GB RAM 32.0 · Disk 45.5 | 30.8 GB RAM 46.2 · Disk 71.5 | 152.1 GB RAM 228.2 · Disk 110.5 |
| GGUF4bit | 11.0 GB RAM 32.0 · Disk 45.6 | 30.9 GB RAM 46.4 · Disk 71.6 | 152.3 GB RAM 228.4 · Disk 110.6 |
GPU Recommendations
1× H100 SXM
80 GB VRAM
≈ 819.1 tok/s
Effective production throughput ≈ 573.4 tok/s
score 0.5919
2× Tesla V100
32 GB VRAM
≈ 0 tok/s (N/A — benchmark unavailable)
Effective production throughput ≈ 0 tok/s
score 0.5933
8× H200 SXM
1128 GB VRAM
≈ 3755.5 tok/s
Effective production throughput ≈ 2628.9 tok/s
score 0.2197
1× RTX A6000
48 GB VRAM
≈ 187.8 tok/s
Effective production throughput ≈ 131.5 tok/s
score 0.6617
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
RTX A6000
48 GB VRAM
$210
/mo · vast
RTX 5090
32 GB VRAM
$236
/mo · vast
L40S
48 GB VRAM
$341
/mo · vast
RTX PRO 6000 WS
48 GB VRAM
$482
/mo · vast
RTX PRO 6000 S
48 GB VRAM
$633
/mo · vast