Image generation (Stable Diffusion, Flux, …).
Diffusion training is workstation-friendly; Flux fine-tuning fits on a single 24GB consumer GPU. SDXL inference runs comfortably on 8–12GB VRAM.
Top GPUs for this workload.
Ranked by suitability — higher fitness scores mean the card handles this workload more comfortably.
| GPU | Tier | VRAM | Fit | |
|---|---|---|---|---|
| consumer | 32GB | 90 | Compare → | |
| consumer | 24GB | 90 | Compare → | |
| workstation | 48GB | 80 | Compare → | |
| consumer | 16GB | 75 | Compare → | |
| datacenter | 48GB | 75 | Compare → | |
| workstation | 48GB | 70 | Compare → | |
| consumer | 24GB | 70 | Compare → | |
| workstation | 32GB | 70 | Compare → | |
| datacenter | 40GB | 70 | Compare → | |
| workstation | 24GB | 65 | Compare → | |
| datacenter | 24GB | 60 | Compare → | |
| workstation | 16GB | 55 | Compare → |
Top models for this workload.
FLUX.1 Pro
Black Forest Labs' flagship image-gen model — closed/API.
Stable Diffusion XL
Workhorse open-weight image-gen — 3.5B params, runs anywhere.
FLUX.1 Dev
Open-weight FLUX.1 — non-commercial license.
Stable Diffusion 3.5 Large
Stability AI's latest open-weight image-gen — 8.1B params, MMDiT architecture.
FLUX.1 Schnell
Distilled fast FLUX.1 — Apache-2.0, commercial-friendly.
Stable Diffusion 3.5 Medium
2.5B SD 3.5 — fits on 12 GB consumer GPUs.
Stable Diffusion 1.5
The original viral image-gen model — still searched heavily.