Presets (for Feedback Input)
Loading presets...
Original Style (Your own writing)
Load Style Template (from style.json):
-- Select a style template --
Feedback Input (Items to be transferred)
Clear Input
Transfer Style
Controls
Perplexity (0-10):
2
Burstiness (0-10):
2
Lexical Diversity (0-10):
2
Academic Style (0-10):
0
Temperature (0.0-2.0):
1.4
Mood (0-Harsh, 5-Neutral, 10-Excited):
5
Sandwich Feedback (0-Source, 10-Max):
0
Remove Articles (a, an, the)
Remove Contractions (e.g., I've -> Ive)
Select Model
qwen3-4b (Local LM Studio)
qwen3-14b (Local LM Studio)
Gemini 1.5 Flash (API)
API Key (Google AI Studio):
AI Detector Sites
Undetectable.ai
Decopy.ai
Phrasly AI
Sapling
Originality.ai
ZeroGPT
Detecting-AI
Scispace
StealthGPT
GPTZero
Rule-Based Rewriter
Note:
This is a proof-of-concept. "WordNet" is a tiny hardcoded subset. Rules are highly simplified and may produce awkward or incorrect results.
Select Rules to Apply:
Select All
Deselect All
Feedback Input ยป Generated Text
Apply to Generated Text (In-place)
Execution Log:
Log will appear here...
Reasoning Process
Waiting for action...
Generated Text
Copy Generated Text