Alibaba Qwen Releases Five Qwen-Audio-3.1 Models with Massive Price Cuts
On September 23, Alibaba's Qwen team released the Qwen-Audio-3.1 family—five models spanning understanding, generation, interaction, and creation—alongside steep price cuts that became the headline of the launch.
Confirmed
- The full audio stack: upgraded ASR, TTS, and Realtime models, plus new ASR-Next (next-generation audio understanding) and TTS-Next (audio creation)
- Major price cuts: ASR down 95%, Realtime down 85%, TTS down about 70%; Qwen-Audio-3.1-ASR-Flash is listed at 0.8 RMB per million tokens
- ASR supports 30 languages plus 16 Chinese dialects, with native text polishing that removes filler words and repetitions
- Per @aigclink, ASR, TTS, TTS-Next, and Realtime APIs are live; ASR-Next API is coming soon
Why it matters
- Ultra-low speech API pricing (ASR down 95%) lowers barriers for voice app developers and could trigger a new price war in the speech market
- The five-model full stack signals Alibaba's aim to compete head-on across the entire voice capability chain
2026-09-23 ~ 2026-09-23 · 6 related posts
Primary sources
- [source] Alibaba's Qwen-Audio-3.1 ships five voice models, cuts ASR price 95% — 千问大模型 · 2026-09-23
- [source] Qwen-Audio-3.1 models go live: ASR-Flash priced at ¥0.8/M input tokens — aigclink · 2026-09-23
- [source] Alibaba's Qwen Launches Five-Model Audio Stack, Slashes TTS 70% and ASR up to 95% — Alibaba_Qwen · 2026-09-23
3 near-duplicate retellings: aigclink · airesearch12 · 智东西