Aesthetic Styler (Advanced)
The same 55 styles, six outputs for SDXL splitting
- text_positive_g
- text_positive_l
- text_positive
- text_negative_g
- text_negative_l
- text_negative
Same 55 art-movement presets as the plain Aesthetic Styler, but the Advanced variant trades one prompt box for two and hands you six outputs. AestheticStylerAdvanced from aegis72/comfyui-styles-all is built for people who want to drive SDXL's two text encoders separately: a G (global) string for the OpenCLIP-G encoder, an L (local) string for CLIP-L, plus a combined output and routed negatives.
The inputs
text_positive_gandtext_positive_l- two positive boxes, meant to feed the two SDXL encoders.text_negative- your negative; the style's negative is appended.Aesthetic- the 55-option dropdown (Art Deco Glamour,Neon Dreams,Zen Garden, …).negative_prompt_to-Both(default),G only, orL only: which encoder(s) receive the negative.log_prompt- on by default; prints the full before/after to the console.
Outputs: text_positive_g, text_positive_l, text_positive (combined), text_negative_g, text_negative_l, text_negative.
A quirk worth knowing before you wire it up
The aesthetic templates are single-part wrappers - no . separator - so the G/L machinery has less to split than it wants. And if you actually run the code, the two positive boxes trade places: your text_positive_l text is what gets styled and exits as text_positive_g, while text_positive_g passes through unstyled to text_positive_l.
Practical guidance:
- Fill
text_positive_lwith your real prompt and take the combinedtext_positiveoutput into a normalCLIPTextEncode- the safest path. - On SDXL with
CLIPTextEncodeSDXL, you can use the combinedtext_positive/text_negativestraight into its text_g/text_l inputs. - If you only want "one box, one dropdown," the non-Advanced Aesthetic Styler is the less confusing node.
Installing it
ComfyUI Manager (search ComfyUI-styles-all), or:
cd ComfyUI/custom_nodes
git clone https://github.com/aegis72/comfyui-styles-all
Then restart ComfyUI. No dependencies, no models.
Where people get burned
- The swap above is the top one - expect your styled output to look "wrong" until you account for it.
- Restart after editing JSON in
data/Aesthetic/for new menu options to appear. - SDXL-era styling - weaker on Flux and LLM-encoded checkpoints, same as every node in this pack.
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| text_positive_g | STRING | — | |
| text_positive_l | STRING | — | |
| text_negative | STRING | — | |
| Aesthetic | COMBO | 55 options: None, Artistic Exploration, Art Deco Glamour, Artistic Opulence, Artistic Whimsy, Baroque Opulence, +49 | |
| negative_prompt_to | COMBO | Both | 3 options: Both, G only, L only |
| log_prompt | BOOLEAN | true | — |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| text_positive_g | STRING | — |
| text_positive_l | STRING | — |
| text_positive | STRING | — |
| text_negative_g | STRING | — |
| text_negative_l | STRING | — |
| text_negative | STRING | — |