Product Launchttsvoice cloningmultilingual speechopen weight
Qwen3-TTS Delivers High-Fidelity Multilingual Speech For Real-Time Applications
7.1
Prompt Engineering's new guide describes Qwen3-TTS, an open-weight text-to-speech model that offers voice cloning, custom voice design, and multilingual support for up to 10 languages. It ships in two sizes — 1.7 billion and 6 billion parameters — optimized for edge deployment, with low-latency streaming and resource needs of about 3–4 GB GPU VRAM; the guide notes variability and multi-model hardware limits.
Key Points
- 1Offers voice cloning and custom-voice generation with 1.7B and 6B parameter models.
- 2Enables low-latency multilingual output (up to 10 languages), suitable for real-time applications.
- 3Require allocating 3–4 GB GPU VRAM per model; plan for runs to address output variability.
Scoring Rationale
Practical, useable release with detailed specs, but limited novelty compared with existing high-fidelity TTS solutions.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
