TREE NEWS update: Alibaba’s Qwen released the Qwen-Audio-3.1 series of speech models, upgrading its ASR, TTS and Realtime voice interaction models and adding the new Qwen-Audio-3.1-TTS-Next audio creation model and Qwen-Audio-3.1-ASR-Next audio understanding model. Five new speech models were launched together, covering understanding, generation, interaction and creation. Prices across the Qwen-Audio line were cut, with TTS down about 70%, Realtime about 85% and ASR 95%.
Alibaba’s Qwen Releases Qwen-Audio-3.1, Cuts Speech Model Prices by Up to 95%
AI take
The headline number is the price cut, but the more consequential move is the breadth of the release: understanding, generation, interaction and creation models shipped together, which turns speech from a single API call into a stack. That matters most for developers building voice agents, where ASR and TTS costs have been a real constraint on always-on interaction. Whether rivals match these prices, and whether cheap speech plus capable models pulls more builders into voice-first products, is the open question.
Generated by AI for reference only.
Share on WeChat
Open WeChat → Scan → then tap "…" to send to a chat or Moments.
Tap "…" in the top-right corner to send to a chat or share to Moments.
Related News
13m ago
Wrtn Technologies Picks Mirae Asset Securities to Lead IPO
13m ago
Nvidia-backed Firmus in talks for $10B AI chip financing
45m ago
Doubao Work Adds ‘Goal’ and ‘Plan’ Modes to Task Feature
46m ago
ByteDance’s Doubao cuts dialogue teams as focus shifts to monetization
1h ago
Firecrawl Raises $75M Series B, Launches Alexandria Data Platform for AI Agents
1h ago
Chengdu and China Telecom Launch City’s First Token Factory Platform