Pixtral 12B
mistralai/Pixtral-12B-2409
Parameters
12B
Context
128K
Downloads
7.7K
Likes
694
Architecture
transformer
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 31.9 GB RAM 47.8 · Disk 73.9 | 35.3 GB RAM 53.0 · Disk 99.9 | 52.5 GB RAM 78.7 · Disk 138.9 |
| FP88bit | 17.7 GB RAM 32.0 · Disk 56.4 | 20.5 GB RAM 32.0 · Disk 82.4 | 37.0 GB RAM 55.5 · Disk 121.4 |
| INT88bit | 17.7 GB RAM 32.0 · Disk 56.4 | 20.5 GB RAM 32.0 · Disk 82.4 | 37.0 GB RAM 55.5 · Disk 121.4 |
| AWQ4bit | 10.9 GB RAM 32.0 · Disk 48.0 | 13.4 GB RAM 32.0 · Disk 74.0 | 29.6 GB RAM 44.4 · Disk 113.0 |
| GPTQ4bit | 11.0 GB RAM 32.0 · Disk 48.2 | 13.5 GB RAM 32.0 · Disk 74.2 | 29.7 GB RAM 44.6 · Disk 113.2 |
| GGUF4bit | 11.2 GB RAM 32.0 · Disk 48.4 | 13.7 GB RAM 32.0 · Disk 74.4 | 29.9 GB RAM 44.8 · Disk 113.4 |
GPU Recommendations
1× RTX 5090
32 GB VRAM
≈ 309 tok/s
Effective production throughput ≈ 216.3 tok/s
score 0.6079
1× RTX 3090
24 GB VRAM
≈ 161.4 tok/s
Effective production throughput ≈ 113 tok/s
score 0.6439
8× H100 SXM
640 GB VRAM
≈ 1848.3 tok/s
Effective production throughput ≈ 1293.8 tok/s
score 0.2026
1× RTX 4090
24 GB VRAM
≈ 173.8 tok/s
Effective production throughput ≈ 121.7 tok/s
score 0.6424
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
RTX 3090
24 GB VRAM
$89.2
/mo · vast
Tesla V100
16 GB VRAM
$128
/mo · vast
RTX 4070S Ti
16 GB VRAM
$138
/mo · vast
RTX 4080S
16 GB VRAM
$157
/mo · vast
A10
24 GB VRAM
$171
/mo · vast