Can the RTX 5090 32 GB run Chroma1-HD?
Yes. 16-bit fits on the RTX 5090 32 GB with 11.9 GB to spare.
16-bit (17.8 GB) needs about 20.1 GB in use, so 11.9 GB of the 32 GB stays free: room for a larger image, a LoRA or a ControlNet.
01What to download
| Part | File | Folder | Size | |
|---|---|---|---|---|
| Model | Chroma1-HD.safetensors 16-bit · lodestones/Chroma1-HD | models/diffusion_models | 17.8 GB | Download → |
| Text encoder | t5xxl_fp8_e4m3fn.safetensors T5-XXL FP8 · comfyanonymous/flux_text_encoders | models/text_encoders | 4.9 GB | Download → |
| VAE | ae.safetensors FLUX.1 VAE (ae) · lodestones/Chroma | models/vae | 0.3 GB | Download → |
| Total download · keep about the same free on disk | 23.0 GB | |||
Alternatives: T5-XXL FP16 (9.8 GB). Sizes from Hugging Face, 2026-10-01.
Free ComfyUI workflow, tested on my RTX 5060 Ti: FP8 workflow (.json). Drag the file into ComfyUI. All workflows →
02Every Chroma1-HD file on 32 GB
| File | Size | Needed | On this card | Quality | Download |
|---|---|---|---|---|---|
| 16-bit ← SAFETENSORS · lodestones | 17.8 GB | 20.1 GB | fits · 11.9 GB spare | the original weights | Hugging Face → |
| FP8 SAFETENSORS · Comfy-Org | 9.2 GB | 11.5 GB | fits · 20.5 GB spare | practically identical to the original | Hugging Face → |
| Q8_0 GGUF · silveroxides | 9.7 GB | 12.0 GB | fits · 20.0 GB spare | practically identical to the original | Hugging Face → |
| Q6_K GGUF · silveroxides | 7.7 GB | 10.0 GB | fits · 22.0 GB spare | very close to the original | Hugging Face → |
| Q5_K_M GGUF · silveroxides | 6.7 GB | 9.0 GB | fits · 23.0 GB spare | close; small differences in fine detail | Hugging Face → |
| Q4_K_M GGUF · silveroxides | 5.6 GB | 7.9 GB | fits · 24.1 GB spare | good; some loss in fine detail and text | Hugging Face → |
| Q3_K_S GGUF · silveroxides | 4.3 GB | 6.6 GB | fits · 25.4 GB spare | noticeable loss of detail | Hugging Face → |
| Q2_K GGUF · silveroxides | 3.4 GB | 5.7 GB | fits · 26.3 GB spare | heavy loss; a last resort | Hugging Face → |
“Needed” adds 1.5 GB of working memory for Chroma1-HD and a 0.8 GB reserve. Sizes 2026-10-01.
03Measured and reported results
Nobody has sent measured numbers for this pair yet. If you run Chroma1-HD on a RTX 5090 32 GB, send your time per image and peak VRAM and it will appear here, credited.
04About the RTX 5090 32 GB
The RTX 5090 32 GB is a Blackwell GPU with FP8 and FP4 hardware: ComfyUI computes Comfy-Org's FP8 files natively here, and NVFP4 files (where a model offers them) are faster still.
32 GB GDDR7 · 512-bit · 1792 GB/s · Blackwell. Everything that runs on it →
05Questions
How much VRAM does Chroma1-HD need?
Around 5.7 GB with the smallest file (Q2_K) and 11.5 GB with an 8-bit file (FP8). The full 16-bit file needs about 20.1 GB.
Which Chroma1-HD file should I download for the RTX 5090 32 GB?
16-bit (17.8 GB) from lodestones/Chroma1-HD.
FP8 or Q8_0 for Chroma1-HD on the RTX 5090 32 GB?
FP8 (9.2 GB): the RTX 5090 32 GB computes it natively, while Q8_0 (9.7 GB) is unpacked on the fly.
How much do I need to download for Chroma1-HD on the RTX 5090 32 GB?
23.0 GB: Chroma1-HD.safetensors, t5xxl_fp8_e4m3fn.safetensors, ae.safetensors.