Open Standard ยท Version 1.0
Version 1.0 ยท March 30, 2026 ID: QARI-QAEF-2026-001 Open Standard

Quranic AI Evaluation Framework

QAEF is QariAI's implementation of an open, standardized methodology for evaluating AI Quranic recitation systems โ€” the first of its kind.

Technical specification on GitHub โ†’

Definition: Quranic recitation evaluation is the measurement of phonetic accuracy, rule compliance, temporal precision, feedback quality, and robustness across speakers.

Five Evaluation Dimensions
1

Phoneme Accuracy

QAEF analyzes recitation at the phoneme level, identifying errors in articulation points (makharij) and phonetic characteristics (sifaat).

ุต vs ุณ confusion ุญ vs ู‡ confusion Incorrect makhraj
2

Tajweed Rule Detection

The system detects application of Tajweed rules in real time and classifies errors by rule type, enabling precise correction.

Ghunnah Idgham Ikhfa Qalqalah Madd Lam Shamsiyyah
3

Timing & Prosody

QAEF evaluates temporal aspects of recitation to ensure alignment with proper recitation structure.

Madd duration (2, 4, 6 counts) Waqf placement Rhythm & flow
4

Feedback Quality

QAEF generates structured feedback designed to be pedagogically actionable โ€” not just flagging errors, but guiding correction.

Accurate Specific Actionable
5

Demographic Robustness

The system is designed to perform consistently regardless of speaker background.

Arab accents South Asian accents Western accents All age groups
QAEF vs Traditional Learning
Dimension Traditional Learning QAEF
Feedback latencyDelayedImmediate
Practice frequencyLimitedUnlimited
ConsistencyVariableStandardized
Error specificityInstructor-dependentRule-based

QAEF complements traditional learning โ€” it enables high-frequency structured practice, not a replacement for scholarship.

Limitations

QAEF is a support system, not a replacement for human scholarship.

See QAEF in action

QariAI applies QAEF to your recitation in real time โ€” identify your weak rules and improve systematically.

Try QariAI free