# Beyond the Benchmark: Conditioning, Domains and Failure Modes of Qwen3-TTS

arXiv: https://arxiv.org/abs/2610.23678 · HTML: https://mlsanity.com/paper/2610.23678/

## Key facts

- No citing papers in the map yet (data through 19 Oct 2026).
- Placed in Vocoders and speech synthesis, one of 11 directions in Speech and audio language models.
- Builds on 22 of its 33 references that are in the map.
- First posted on arXiv 16 Oct 2026.

## Placed in

- Speech and audio language models / [Vocoders and speech synthesis](https://mlsanity.com/speech/vocoders-and-speech-synthesis/)

## Builds on (most cited first, 22 of 22)

- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers](https://mlsanity.com/paper/2301.02111/) (2023)
- [F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching](https://mlsanity.com/paper/2410.06885/) (2024)
- [NaturalSpeech 3: Zero-Shot Speech Synthesis with Factorized Codec and Diffusion Models](https://mlsanity.com/paper/2403.03100/) (2024)
- [DNSMOS P.835: A Non-Intrusive Perceptual Objective Speech Quality Metric to Evaluate Noise Suppressors](https://mlsanity.com/paper/2110.01763/) (2021)
- [emotion2vec: Self-Supervised Pre-Training for Speech Emotion Representation](https://mlsanity.com/paper/2312.15185/) (2023)
- [XTTS: a Massively Multilingual Zero-Shot Text-to-Speech Model](https://mlsanity.com/paper/2406.04904/) (2024)
- [AASIST: Audio Anti-Spoofing using Integrated Spectro-Temporal Graph Attention Networks](https://mlsanity.com/paper/2110.01200/) (2021)
- [ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection](https://mlsanity.com/paper/1904.05441/) (2019)
- [ASVspoof 2021: accelerating progress in spoofed and deepfake speech detection](https://mlsanity.com/paper/2109.00537/) (2021)
- [Wespeaker: A Research and Production oriented Speaker Embedding Learning Toolkit](https://mlsanity.com/paper/2210.17016/) (2022)
- [VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizers](https://mlsanity.com/paper/2406.05370/) (2024)
- [Qwen3-TTS Technical Report](https://mlsanity.com/paper/2601.15621/) (2026)
- [TorchAudio-Squim: Reference-less Speech Quality and Intelligibility measures in TorchAudio](https://mlsanity.com/paper/2304.01448/) (2023)
- [ELLA-V: Stable Neural Codec Language Modeling with Alignment-guided Sequence Reordering](https://mlsanity.com/paper/2401.07333/) (2024)
- [SpeechBERTScore: Reference-Aware Automatic Evaluation of Speech Generation Leveraging NLP Evaluation Metrics](https://mlsanity.com/paper/2401.16812/) (2024)
- [Audio-Aware Large Language Models as Judges for Speaking Styles](https://mlsanity.com/paper/2506.05984/) (2025)
- [ATCO2 corpus: A Large-Scale Dataset for Research on Automatic Speech Recognition and Natural Language Understanding of Air Traffic Control Communications](https://mlsanity.com/paper/2211.04054/) (2022)
- [Warning: Humans Cannot Reliably Detect Speech Deepfakes](https://mlsanity.com/paper/2301.07829/) (2023)
- [SpeechJudge: Towards Human-Level Judgment for Speech Naturalness](https://mlsanity.com/paper/2511.07931/) (2025)
- [People are poorly equipped to detect AI-powered voice clones](https://mlsanity.com/paper/2410.03791/) (2024)
- [The State Of TTS: A Case Study with Human Fooling Rates](https://mlsanity.com/paper/2508.04179/) (2025)

## Abstract (arXiv)

Qwen3-TTS reports state-of-the-art zero-shot voice cloning, but its technical report leaves open how the model should be conditioned, how it behaves outside read speech, and how it fails. We characterise its two cloning interfaces, in-context learning (ICL) and X-Vector, against Chatterbox, XTTS-v2 and F5-TTS on read, meeting, emotional and air-traffic-control (ATC) speech, using eighteen objective measures, an LLM judge, a spoofing countermeasure and a listening test. Neither interface dominates: ICL clones timbre more closely from correctly transcribed references, X-Vector clones timing, and ICL degrades sharply when a short reference transcript is wrong. Domain explains four times more variance than system choice. Qwen3-TTS leads on intelligibility and pitch register but trails on prosody, and the objective measures rank systems differently from listeners. An off-the-shelf countermeasure flags its clones no more often than genuine recordings, and on ATC radio it fails for every system. Code and audio are released.