PaliGemma 3B
google/paligemma-3b-pt-224
Parameters
3B
Context
8K
Downloads
171.6K
Likes
540
Architecture
transformer
GPU Requirements
VRAM / RAM / disk estimates by quantization and usage scenario
| Quantization | Minimum | Recommended | Production |
|---|---|---|---|
| FP1616bit | 10.6 GB RAM 32.0 · Disk 47.7 | 13.1 GB RAM 32.0 · Disk 73.7 | 29.3 GB RAM 44.0 · Disk 112.7 |
| FP88bit | 7.1 GB RAM 32.0 · Disk 43.4 | 9.5 GB RAM 32.0 · Disk 69.4 | 25.5 GB RAM 38.2 · Disk 108.4 |
| INT88bit | 7.1 GB RAM 32.0 · Disk 43.4 | 9.5 GB RAM 32.0 · Disk 69.4 | 25.5 GB RAM 38.2 · Disk 108.4 |
| AWQ4bit | 5.4 GB RAM 32.0 · Disk 41.3 | 7.7 GB RAM 32.0 · Disk 67.3 | 23.6 GB RAM 35.4 · Disk 106.3 |
| GPTQ4bit | 5.4 GB RAM 32.0 · Disk 41.3 | 7.7 GB RAM 32.0 · Disk 67.3 | 23.6 GB RAM 35.4 · Disk 106.3 |
| GGUF4bit | 5.5 GB RAM 32.0 · Disk 41.3 | 7.7 GB RAM 32.0 · Disk 67.3 | 23.7 GB RAM 35.5 · Disk 106.3 |
GPU Recommendations
1× RTX 3080 Ti
12 GB VRAM
≈ 629 tok/s
Effective production throughput ≈ 440.3 tok/s
score 0.6655
1× Tesla V100
16 GB VRAM
≈ 0 tok/s (N/A — benchmark unavailable)
Effective production throughput ≈ 0 tok/s
score 0.617
8× RTX 5090
256 GB VRAM
≈ 3954.8 tok/s
Effective production throughput ≈ 2768.4 tok/s
score 0.3522
1× RTX 4060 Ti
8 GB VRAM
≈ 198.6 tok/s
Effective production throughput ≈ 139 tok/s
score 0.6375
💬 Community — real deployment experience
0 Articles · 0 Benchmarks
GPUs that fit (single card)
Tesla V100
16 GB VRAM
$39.1
/mo · vast
RTX 4060 Ti
8 GB VRAM
$59.5
/mo · vast
RTX 4070S Ti
16 GB VRAM
$68.5
/mo · vast
RTX 5060 Ti
8 GB VRAM
$68.7
/mo · vast
RTX 3080 Ti
12 GB VRAM
$69.2
/mo · vast