Loanword Accent Predictor

Predicted Accent Core (Downstep)
Type in Hiragana or Katakana
The "Antepenultimate" Rule
Core Principle: For Japanese loanwords (Gairaigo), the accent nucleus typically falls on the Antepenultimate Mora (3rd from the end).
However, if that mora is "Special" (cannot bear accent), it shifts to the Pre-antepenultimate (4th from end).
Scenario Logic Example
Standard Case
Antepenult is normal		 Target: -3 Mora
Result: Accent Here 		カメラ (Camera)
Morae: ka-me-ra
Target (-3): ka
Output: カメラ
Conflict Case 1
Antepenult is "N" (ん)		 Target: N (Skip)
Shift: -4 Mora 		カレンダー (Calendar)
Morae: ka-re-n-da-a
Target (-3): n NO
Shift (-4): re
Output: カレンダー
Conflict Case 2
Antepenult is "Long Vowel" (ー)		 Target: ー (Skip)
Shift: -4 Mora 		コンピューター
Morae: ...pyu-u-ta-a
Target (-3): u NO
Shift (-4): pyu
Output: コンピューター
Conflict Case 3
Antepenult is "Small Tsu" (ッ)		 Target: ッ (Skip)
Shift: -4 Mora 		インデックス
Morae: ...i-n-de-k-ku-su
Target (-3): k NO
Shift (-4): de
Output: インデックス
* Note: This is a predictive heuristic. Actual accent may vary based on word origin and length.