Can the RTX 4060 Ti 8 GB run ERNIE-Image (Turbo)?
Not at full quality. The 16-bit and 8-bit files are too big for the RTX 4060 Ti 8 GB; the largest that fits is Q4_K_M (5.0 GB).
Q4_K_M (5.0 GB) fits with only 0.7 GB to spare (about 7.3 GB of 8 GB). Quality: good; some loss in fine detail and text. For better quality, Q5_K_M (5.9 GB) also runs, with about 0.2 GB spilling into system RAM - a little slower, still practical.
01What to download
| Part | File | Folder | Size | |
|---|---|---|---|---|
| Model | ernie-image-turbo-Q4_K_M.gguf Q4_K_M · unsloth/ERNIE-Image-Turbo-GGUF | models/unet | 5.0 GB | Download → |
| Text encoder | ministral-3-3b.safetensors Ministral 3 3B · Comfy-Org/ERNIE-Image | models/text_encoders | 7.7 GB | Download → |
| VAE | flux2-vae.safetensors FLUX.2 VAE · Comfy-Org/ERNIE-Image | models/vae | 0.3 GB | Download → |
| Total download · keep about the same free on disk | 13.1 GB | |||
Only for some workflows: ERNIE prompt enhancer (optional) (6.9 GB). Sizes from Hugging Face, 2026-10-01.
02Every ERNIE-Image file on 8 GB
| File | Size | Needed | On this card | Quality | Download |
|---|---|---|---|---|---|
| 16-bit SAFETENSORS · Comfy-Org | 16.1 GB | 18.4 GB | 10.4 GB too big | the original weights | Hugging Face → |
| Q8_0 GGUF · unsloth | 8.7 GB | 11.0 GB | 3.0 GB too big | practically identical to the original | Hugging Face → |
| Q6_K GGUF · unsloth | 6.8 GB | 9.1 GB | spills 1.1 GB | very close to the original | Hugging Face → |
| Q5_K_M GGUF · unsloth | 5.9 GB | 8.2 GB | spills 0.2 GB | close; small differences in fine detail | Hugging Face → |
| Q4_K_M ← GGUF · unsloth | 5.0 GB | 7.3 GB | fits · 0.7 GB spare | good; some loss in fine detail and text | Hugging Face → |
| Q3_K_M GGUF · unsloth | 3.9 GB | 6.2 GB | fits · 1.8 GB spare | noticeable loss of detail | Hugging Face → |
| Q2_K GGUF · unsloth | 3.2 GB | 5.5 GB | fits · 2.5 GB spare | heavy loss; a last resort | Hugging Face → |
“Needed” adds 1.5 GB of working memory for ERNIE-Image and a 0.8 GB reserve. Sizes 2026-10-01.
03The text encoder
At 7.7 GB even the smallest Ministral 3 3B file is too big for 8 GB, so it runs from system RAM. That slows down reading the prompt, not the picture. Sizes: 16-bit 7.7 GB (an optional prompt-enhancer LLM (6.9 GB) can be skipped). Why the encoder does not count against VRAM →
04Measured and reported results
Nobody has sent measured numbers for this pair yet. If you run ERNIE-Image on a RTX 4060 Ti 8 GB, send your time per image and peak VRAM and it will appear here, credited.
05With more VRAM
With 12 GB you could run ERNIE-Image at 8-bit or better (Q8_0, 8.7 GB) with no offloading: for example on the RTX 5070 12 GB, RTX 4070 Super 12 GB, RTX 4070 12 GB.
What changes from the RTX 4060 Ti 8 GB to the RTX 5060 Ti 16 GB →
06About the RTX 4060 Ti 8 GB
The RTX 4060 Ti 8 GB is an Ada Lovelace GPU with hardware FP8, so ComfyUI can compute Comfy-Org's FP8 files natively: small and fast. (Plain FP8 files use FP8 maths with the --fast fp8_matrix_mult option.)
8 GB GDDR6 · 128-bit · 288 GB/s · Ada Lovelace. Everything that runs on it →
07Questions
How much VRAM does ERNIE-Image need?
Around 5.5 GB with the smallest file (Q2_K) and 11.0 GB with an 8-bit file (Q8_0). The full 16-bit file needs about 18.4 GB.
Which ERNIE-Image file should I download for the RTX 4060 Ti 8 GB?
Q4_K_M (5.0 GB) from unsloth/ERNIE-Image-Turbo-GGUF. Load it with Unet Loader (GGUF) from the ComfyUI-GGUF node pack.
How much do I need to download for ERNIE-Image on the RTX 4060 Ti 8 GB?
13.1 GB: ernie-image-turbo-Q4_K_M.gguf, ministral-3-3b.safetensors, flux2-vae.safetensors.