Can the RTX 5090 32 GB run Ideogram 4?
Yes. FP8 fits on the RTX 5090 32 GB with 11.1 GB to spare.
FP8 (9.3 GB (×2, both loaded)) needs about 20.9 GB in use, so 11.1 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 | ideogram4_fp8_scaled.safetensors main model FP8 · Comfy-Org/Ideogram-4 | models/diffusion_models | 9.3 GB | Download → |
| Model | ideogram4_unconditional_fp8_scaled.safetensors unconditional model, loaded together with the main one FP8 · Comfy-Org/Ideogram-4 | models/diffusion_models | 9.3 GB | Download → |
| Text encoder | qwen3vl_8b_fp8_scaled.safetensors Qwen3-VL 8B FP8 · Comfy-Org/Ideogram-4 | models/text_encoders | 10.6 GB | Download → |
| VAE | flux2-vae.safetensors FLUX.2 VAE · Comfy-Org/Ideogram-4 | models/vae | 0.3 GB | Download → |
| Total download · keep about the same free on disk | 29.5 GB | |||
Alternatives: Qwen3-VL 8B NVFP4 (6.3 GB). FP4 files are listed only as alternatives unless the GPU is an RTX 50 (Blackwell). Sizes from Hugging Face, 2026-10-01.
System RAM: 48 GB or more. These files (both noise models) take about 29.5 GB of RAM while loaded, 35.5 GB with Windows and ComfyUI. What happens with less. calculated
02Every Ideogram 4 file on 32 GB
| File | Size | Needed | On this card | Quality | Download |
|---|---|---|---|---|---|
| FP8 ← SAFETENSORS · Comfy-Org | 9.3 GB ×2 | 20.9 GB | fits · 11.1 GB spare | practically identical to the original | Hugging Face → |
| INT8 SAFETENSORS · Comfy-Org | 9.6 GB ×2 | 21.5 GB | fits · 10.5 GB spare | practically identical to the original | Hugging Face → |
| Q8_0 GGUF · molbal | 10.1 GB ×2 | 22.6 GB | fits · 9.4 GB spare | practically identical to the original | Hugging Face → |
| Q5_1 GGUF · molbal | 7.3 GB ×2 | 17.0 GB | fits · 15.0 GB spare | close; small differences in fine detail | Hugging Face → |
| Q4_1 GGUF · molbal | 6.2 GB ×2 | 14.7 GB | fits · 17.3 GB spare | good; some loss in fine detail and text | Hugging Face → |
“Needed” adds 1.5 GB of working memory for Ideogram 4 and a 0.8 GB reserve. Sizes 2026-10-01. Needs two transformer files loaded together (main + unconditional). GGUF files need the molbal fork of ComfyUI-GGUF. No 16-bit weights were released.
03Measured and reported results
Nobody has sent measured numbers for this pair yet. If you run Ideogram 4 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 Ideogram 4 need?
Around 14.7 GB with the smallest file (Q4_1) and 20.9 GB with an 8-bit file (FP8).
Which Ideogram 4 file should I download for the RTX 5090 32 GB?
FP8 (9.3 GB) from Comfy-Org/Ideogram-4.
FP8 or Q8_0 for Ideogram 4 on the RTX 5090 32 GB?
FP8 (9.3 GB): the RTX 5090 32 GB computes it natively, while Q8_0 (10.1 GB) is unpacked on the fly.
How much do I need to download for Ideogram 4 on the RTX 5090 32 GB?
29.5 GB: ideogram4_fp8_scaled.safetensors, ideogram4_unconditional_fp8_scaled.safetensors, qwen3vl_8b_fp8_scaled.safetensors, flux2-vae.safetensors.