Phase 2: Dialectal Arabic ASR System
Using ArTST v3 (ACL 2025). This model handles MSA and 17 regional dialects.
Upload/Record Arabic Speech
Drop Audio Here
- or -
Click to Upload
Ground Truth (Optional - for Evaluation)
Clear
Submit
Model Transcription (ArTST v3)
Performance Metrics
Share via Link