Qwen3-TTS Hits SageMaker JumpStart: Zero-Shot Voice Cloning From Seconds of Audio
AWS ML Blog · rss · 2026-09-26
Alibaba's Qwen3-TTS-12Hz-1.7B-Base is now on Amazon SageMaker JumpStart, offering zero-shot voice cloning from a few seconds of reference audio across 10 languages with cross-lingual cloning and streaming. The walkthrough covers deploying to a real-time endpoint on a single L4 GPU, including the vLLM memory split (0.45 utilization) needed since talker and code2wav stages share one GPU.
More from Infra
- Moody's Warns Anthropic and OpenAI Carry $2.5T in Off-Balance-Sheet Debt — SumitGup · 2026-09-26
- SemiAnalysis Maps 1,000+ China Datacenters Across 60+ Operators in New AI Infrastructure Model — zephyr_z9 · 2026-09-26
- Everything About Hosting Got Cheaper Except Moderation: LessWrong Burns ~$1M/Year — jd_pressman · 2026-09-26
- SpaceX unveils AI training cluster site: $90B+ invested, 7,500 local jobs, 3.3 GWh Megapacks — elonmusk · 2026-09-26
- OpenAI Has ~1.9 GW of Compute, Wants 30 GW; US Buildout Nears 100 GW — le_james94 · 2026-09-26
- Inference Is the COGS of AI: Gross Margin Is Set by Token Cost and Speed — le_james94 · 2026-09-26