No benchmark data yet
Run make benchmark on a GPU worker, then make benchmark-publish to copy results.json into the site.
Model specifications
Hardware and license requirements from worker manifests. VRAM is for full GPU residency; lower-VRAM cards may use CPU offload at reduced speed.
| Model | Architecture | Parameters | Min VRAM | Resolutions | Steps | Capabilities | License | Commercial | In studio |
|---|---|---|---|---|---|---|---|---|---|
SDXL (default) sdxl-base | SDXL | ~3.5B | 10 GB | 768, 1024 | 10-50 | text to image, image to image | CreativeML Open RAIL++-M | Unrestricted (RAIL use policy) | Yes |
SDXL Fast sdxl-fast | SDXL + Lightning LoRA | ~3.5B | 10 GB | 1024 | 4-8 | text to image | RAIL++-M + ByteDance Lightning | Unrestricted (RAIL use policy) | Yes |
SSD-1B ssd-1b | SDXL distilled | ~1.3B | 8 GB | 768, 1024 | 10-40 | text to image | Apache 2.0 | Unrestricted | Yes |
DreamShaper 8 LCM dreamshaper-lcm | SD 1.5 LCM | ~860M | 6 GB | 512, 768 | 2-15 | text to image | CreativeML Open RAIL-M | Unrestricted (RAIL use policy) | Reference only |
SD Turbo sd-turbo | SD 1.5 distilled | ~860M | 8 GB | 512 | 1-4 | text to image, image to image, realtime | Stability AI Community | ≤ $1M revenue / year | Yes |
SDXL Turbo sdxl-turbo | SDXL distilled | ~3.5B | 10 GB | 512 | 1-4 | text to image, image to image, realtime | Stability AI Community | ≤ $1M revenue / year | Yes |
SDXL Hyper-SD sdxl-hypersd | SDXL + Hyper-SD LoRA | ~3.5B | 10 GB | 1024 | 4-8 | text to image | RAIL++-M base; LoRA: no declared license | Unclear - LoRA license undeclared | Yes |
Segmind VegaRT vega-rt | Segmind-Vega + VegaRT LCM LoRA | ~1.0B | 8 GB | 512, 1024 | 2-8 | text to image, image to image, realtime | Apache 2.0 | Unrestricted | Yes |
SSD-1B + Lightning ssd-1b-lightning | SSD-1B + Lightning LoRA | ~1.3B | 8 GB | 1024 | 4-8 | text to image | Apache 2.0 base + Open RAIL++-M LoRA | RAIL use policy applies | Yes |
Stable Diffusion 3.5 Medium sd35-medium | MMDiT + CLIP-L + OpenCLIP-G + T5-XXL | ~2.5B | 14 GB | 768, 1024 | 20-50 | text to image | Stability AI Community | <= $1M revenue / year | Yes |