Ideogram4 ComfyUI Quantization Acceleration
Ant_6431 · reddit · 2026-07-14
ComfyUI Quantization for Ideogram 4
The post provides a set of ComfyUI model files for Ideogram 4, including ideogram4int8convrot.safetensors, ideogram4unconditionalint8convrot.safetensors, qwen3vl8bfp8scaled.safetensors, and flux2-vae.safetensors. The author offers the default preset V4DEFAULT20, which takes about 51 seconds to render 1MP on a 4070S.
Fast / Instant Versions
Additionally, two lighter versions are provided:
- Ideogram4-fastint8-convrot-simple.safetensors: V4FAST20, no negative prompt, cfg=1, 21 seconds for 1MP
- Ideogram4-instantint8-convrot-simple.safetensors: V4INSTANT8, no negative prompt, cfg=1, 7 seconds for 1MP
The post concludes with the corresponding Hugging Face links for direct downloading and use within ComfyUI.
More from Multimodal
- Reddit user seeks ComfyUI NSFW text-to-image and image-to-video workflows under 20 GB VRAM — hobbyist2020 · 2026-07-22
- Krea 2 users recommend a two-pass Clownshark sampler setup for sharper image details — listopalafoto · 2026-07-22
- Gemini Omni Flash turns a boat cabin into a cave in Flow by Google — chrisfirst · 2026-07-22
- A simple workflow to turn a photo into an image prompt using Gemini, Grok, or GPT Image — harshitagu72595 · 2026-07-22
- A Reddit user proposes a consistency LoRA to keep anime and game scenes visually stable — ThirdWorldBoy21 · 2026-07-22
- Hand-painted figurines run through Seedance look eerily alive — cocktailpeanut · 2026-07-22