Khmer Speech Recognition

Transcribe Khmer speech with seanghay/Qwen3-ASR-0.6B-Khmer-v4, a fine-tune of Qwen3-ASR-0.6B. Long audio is split into ~15s chunks at silences (max 5 min).

This model is for sale. If you need it, please contact seanghay.dev@gmail.com.