kaushiknishchay/ComfyUI-Qwen3-ASR

ComfyUI nodes for Qwen3-ASR (0.6B/1.7B) and ForcedAligner. Supports high-accuracy ASR and language identification for 52 languages/dialects, including 22 Chinese dialects and various English accents. Features word-level timestamps, long audio transcription, and VRAM-optimized inference.

38
/ 100
Emerging

Integrates Qwen3-ForcedAligner as an optional secondary model to generate word-level timestamps through forced alignment rather than direct prediction. Implements chunked audio processing with configurable overlap (default 30s chunks with 2s context) to handle long recordings, plus automatic 16kHz resampling and optional FlashAttention 2 acceleration for reduced VRAM footprint across mixed precision modes (bf16/fp16/fp32).

No Package No Dependents
Maintenance 10 / 25
Adoption 5 / 25
Maturity 9 / 25
Community 14 / 25

How are scores calculated?

Stars

11

Forks

3

Language

Python

License

MIT

Last pushed

Mar 05, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/kaushiknishchay/ComfyUI-Qwen3-ASR"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.