GPU Matcher
Find compatible image generation models for your GPU
Enthusiast/Professional • $1.50-3.00/hour (Vast.ai, RunPod)
GPU Specifications
Tier
Enthusiast/Professional
Cost (Rental)
$1.50-3.00/hour (Vast.ai, RunPod)
Best overall for image generation. Silent, power-efficient for deep learning.
Compatible Models
Flux Dev
Black Forest Labs
Best for this GPU
FP32
Required VRAM
18 GB
Quality Loss
0%
Headroom
6.0 GB free
Full Precision (FP32): Only if VRAM permits. Maximum quality.
Flux Schnell
Black Forest Labs
Best for this GPU
FP32
Required VRAM
14 GB
Quality Loss
0%
Headroom
10.0 GB free
Full Precision (FP32): Only if VRAM permits. Maximum quality.
Stable Diffusion 3.5
Stability AI
Best for this GPU
FP32
Required VRAM
16 GB
Quality Loss
0%
Headroom
8.0 GB free
Full Precision (FP32): Only if VRAM permits. Maximum quality.
Together AI Images
Together AI
Best for this GPU
FP32
Required VRAM
15 GB
Quality Loss
0%
Headroom
9.0 GB free
Full Precision (FP32): Only if VRAM permits. Maximum quality.
Replicate / ComfyUI Workflows
Replicate Inc. (open ecosystem)
Best for this GPU
FP32
Required VRAM
16 GB
Quality Loss
0%
Headroom
8.0 GB free
Full Precision (FP32): Only if VRAM permits. Maximum quality.
💡 Tip: All listed models can run on this GPU. The "Best" quantization shows the highest quality option that fits. Lower quantization levels use less VRAM but with slight quality loss.