High Latency

#7
by wioelm - opened

Hello, the API for Qwen3 ASR Flash API is very slow, (3-4s) and it's not due to geographical location. This makes it unusable for realtime transcription. I was wondering if Alibaba has plans to make this faster, and or release the model so that other companies can host it at a faster speed.

Thank you!

Sign up or log in to comment