Qwen3 TTS is now available on Together AI
Blog post from Together AI
Qwen3 TTS, an open text-to-speech model released under the Apache 2.0 license, is now available on Together AI’s serverless inference and dedicated endpoints for $20 per million characters. It offers nine preset voices that can each speak ten languages, with delivery shaped by plain-language instructions such as “excited and upbeat” while keeping the selected voice consistent; Qwen recommends using each voice in its native language for best quality. Together AI reports first audio in as little as 97 milliseconds and supports streaming through HTTP for complete text or WebSockets for text supplied in pieces, making the model suitable for real-time voice applications. Developers can compare voices and test custom text in VoiceFinder, and generate audio in WAV, MP3, raw PCM, or μ-law formats at supported sample rates of 24 or 8 kHz; Opus, AAC, FLAC, and word-level timestamps are not supported.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| Real-time | 9 | 3,194 | 698 | 199 | -24% |
| Serverless | 3 | 833 | 236 | 92 | -40% |
| Voice AI | 2 | 1,653 | 202 | 45 | -40% |
| LLM | 1 | 3,674 | 823 | 199 | -44% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.