Nodes/ComfyUI-AlchemicaSampler/⚗ Alchemica · Prompt enhancer (text encoder)
ComfyUI Node

⚗ Alchemica · Prompt enhancer (text encoder)

Prompt enhancer that uses Krea 2's own text encoder: no Ollama, it can also read images.

By Nessuno76·Created 2 days ago·Updated 2 days ago· 1
⚗ Alchemica · Prompt enhancer (text encoder)
  • clip
  • immagine
  • scena
  • dettaglio
  • negativo
  • report
◄idea►
◄stilefoto realistica►
◄parole80►
◄creativita0.00►
◄seed0►
Category⚗ AlchemicaMente/sampling

Inputs (7)

NameTypeDefaultDescription
clipCLIPKrea 2's text encoder: it writes the prompt, then encodes it
ideaSTRINGAny language, a few words are enough. Empty + image = describes the image
stileCOMBOfoto realistica8 options: foto realistica, ritratto, still life / laboratorio, paesaggio, architettura, cinematografico, +2
paroleINT8030–200Length of the SCENE (the detail is about one third)
creativitaFLOAT0.000–1.50 = deterministic (same text every run). 0.6-0.9 = variations; change the seed
seedINT00–18446744073709550000Only matters when creativita (creativity) > 0
immagineoptIMAGEVisual reference: the model looks at it (Qwen3-VL is a vision-language model)

Outputs (4)

NameTypeDescription
scenaSTRING—
dettaglioSTRING—
negativoSTRING—
reportSTRING—