Llama 3.2 90B Vision
meta-llama/Llama-3.2-90B-Vision
Parameters
90B
Context
128K
Downloads
113
Likes
134
Architecture
transformer
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 215.6 GB RAM 323.5 · Disk 300.5 | 227.5 GB RAM 341.2 · Disk 326.5 | 252.9 GB RAM 379.4 · Disk 365.5 |
| FP88bit | 109.6 GB RAM 164.4 · Disk 169.8 | 116.6 GB RAM 174.9 · Disk 195.8 | 137.3 GB RAM 205.9 · Disk 234.8 |
| INT88bit | 109.6 GB RAM 164.4 · Disk 169.8 | 116.6 GB RAM 174.9 · Disk 195.8 | 137.3 GB RAM 205.9 · Disk 234.8 |
| AWQ4bit | 58.6 GB RAM 87.9 · Disk 106.8 | 63.3 GB RAM 94.9 · Disk 132.8 | 81.6 GB RAM 122.4 · Disk 171.8 |
| GPTQ4bit | 59.2 GB RAM 88.9 · Disk 107.7 | 64.0 GB RAM 95.9 · Disk 133.7 | 82.3 GB RAM 123.5 · Disk 172.7 |
| GGUF4bit | 60.6 GB RAM 90.9 · Disk 109.3 | 65.3 GB RAM 98.0 · Disk 135.3 | 83.8 GB RAM 125.7 · Disk 174.3 |
GPU Recommendations
1× H200 SXM
141 GB VRAM
≈ 110.4 tok/s
Effective production throughput ≈ 77.3 tok/s
score 0.5907
4× Tesla V100
64 GB VRAM
≈ 0 tok/s (N/A — benchmark unavailable)
Effective production throughput ≈ 0 tok/s
score 0.545
8× H200 SXM
1128 GB VRAM
≈ 353.3 tok/s
Effective production throughput ≈ 247.3 tok/s
score 0.2197
1× A100 80GB
80 GB VRAM
≈ 46.9 tok/s
Effective production throughput ≈ 32.8 tok/s
score 0.6755
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
A100 80GB
80 GB VRAM
$876
/mo · vast
H100 SXM
80 GB VRAM
$1,170
/mo · vast
H200 SXM
141 GB VRAM
$2,555
/mo · lambda
B200
180 GB VRAM
$3,013
/mo · vast
B200 SXM
180 GB VRAM
—