Image models · Stability AI
Stable Cascade
- Backbone
- 5.2B
- Native output
- 1024×1024
- Default steps
- 30
- License
- stable-cascade-nc-community
- Released
- 2024-02
How to run it
Original weights · Hugging Face
Full-precision files from Stability AI, for fine-tuning, vLLM, Transformers, Diffusers or ComfyUI.
huggingface.co/stabilityai/stable-cascade (opens in a new tab)
Official repo · GitHub
Reference code, inference scripts and docs from Stability AI. The weights themselves download from Hugging Face.
github.com/Stability-AI/StableCascade (opens in a new tab)
git clone https://github.com/Stability-AI/StableCascadeMemory needed
Peak GPU memory at the native output size. Offloading moves the text encoder, or the whole pipeline, to system RAM between steps.
| Precision | Everything on GPU | Text encoder offloaded | Sequential offload | Download |
|---|---|---|---|---|
FP16 | 14.0 GB | 12.6 GB | 5.8 GB | Original weights (opens in a new tab) |
FP8 | 8.1 GB | 7.4 GB | 4.0 GB | Original weights (opens in a new tab) |
Components
- Backbone: 5.16B
- VAE: 0.02B
- OpenCLIP bigG: 0.695B
Best hardware for Stable Cascade
The least expensive GPU in our catalog that runs it without offloading is the Intel Arc A380 (launch price $139) at FP8.
| Grade · points | Hardware | Precision | Memory | Time per output | Fit |
|---|---|---|---|---|---|
| 77 ptsGreat (A) · 77 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 | NVIDIA GeForce RTX 4070 SUPER | FP8 | 8.1 GB / 12.0 GB | 8.5 s | Runs well |
| 76 ptsGreat (A) · 76 of 100 points: Model quality 23.4/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 13.4/15 | AMD Radeon RX 7900 XTX | FP16 | 14.0 GB / 24.0 GB | 5.6 s | Runs well |
| 76 ptsGreat (A) · 76 of 100 points: Model quality 23.4/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 13.4/15 | NVIDIA GeForce RTX 3090 | FP16 | 14.0 GB / 24.0 GB | 9.4 s | Runs well |
| 76 ptsGreat (A) · 76 of 100 points: Model quality 23.4/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 13.4/15 | NVIDIA GeForce RTX 4090 | FP16 | 14.0 GB / 24.0 GB | 5.0 s | Runs well |
| 74 ptsGreat (A) · 74 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 11.7/15 | AMD Radeon RX 9070 XT | FP8 | 8.1 GB / 16.0 GB | 6.0 s | Runs well |
| 74 ptsGreat (A) · 74 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 11.7/15 | NVIDIA GeForce RTX 4070 Ti SUPER | FP8 | 8.1 GB / 16.0 GB | 7.2 s | Runs well |
| 74 ptsGreat (A) · 74 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 11.7/15 | NVIDIA GeForce RTX 4080 SUPER | FP8 | 8.1 GB / 16.0 GB | 6.3 s | Runs well |
| 74 ptsGreat (A) · 74 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 11.7/15 | NVIDIA GeForce RTX 5070 Ti | FP8 | 8.1 GB / 16.0 GB | 7.2 s | Runs well |
| 74 ptsGreat (A) · 74 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 11.7/15 | NVIDIA GeForce RTX 5080 | FP8 | 8.1 GB / 16.0 GB | 6.0 s | Runs well |
| 74 ptsGreat (A) · 74 of 100 points: Model quality 22.9/45, Memory fit 25.0/25, Speed 11.1/15, Memory use 15.0/15 | NVIDIA GeForce RTX 3060 12GB | FP8 | 8.1 GB / 12.0 GB | 18 s | Runs well |
Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.