🧪 Prompt Enhancer
Ready
Target model
FLUX (image)
SDXL (image)
WAN video 2.x (video)
OVI (video + speech tags)
x
Add camera/lens
Lighting
Compact
Model
Refresh
Populated from your Ollama server (via config host/port). Filter to narrow, select to use.
x
Live stream
Test connection
Mode
Text → Enhanced
Image → Prompt
Base prompt
Negative prompt (optional)
Upload image
Optional hint
Uses your selected model. If none chosen, falls back to a vision model from config.
Generate
Copy
Download
Enhanced output
Ready.
Debug / errors
What shows here
Connection probe status and version.
Streaming progress: tokens as they arrive.
Final stats from Ollama, if provided.
Behavior by target
FLUX/SDXL:
one clean paragraph; adds camera, lighting, composition; optional “Negative:” line.
WAN video:
single continuous scene; gentle camera motion.
OVI:
visual paragraph, then
Name says <S>…<E>
plus one
<AUDCAP>
.
Image mode:
uploads as raw base64 to your selected model.
Made for ai.kj6ywd.net. Panels collapse now, like my faith in UI libraries.