Text To Speech

Text To Speech
- Survey
- TTS
- Voco
- Emotion
- VITS
- Efficient
- Projects
- Multilingual
- Evaluation
- Misc

Survey

TTS

Qwen3-TTS-Flash: Multi-timbre & Multi-lingual & Multi-dialect Speech Synthesis.
Autoregressive transformer models have proven highly effective for synthesizing speech from text. 🤗
🌟 MiniMax-Speech: Intrinsic Zero-Shot Text-to-Speech with a Learnable Speaker Encoder, arXiv, 2505.07916, arxiv, pdf, cication: -1

Bowen Zhang, Congchao Guo, Geng Yang, ..., Yuan Lu, Yucen He · (minimax-ai.github)
F5R-TTS: Improving Flow-Matching based Text-to-Speech with Group Relative Policy Optimization, arXiv, 2504.02407, arxiv, pdf, cication: -1

Xiaohui Sun, Ruitong Xiao, Jianye Mo, ..., Qun Yu, Baoxun Wang · (frontierlabs.github)
Emilia: An Extensive, Multilingual, and Diverse Speech Dataset for Large-Scale Speech Generation 🤗
Hume AI dropped a text-to-speech LLM that understands emotional context 𝕏
Long-Form Speech Generation with Spoken Language Models, arXiv, 2412.18603, arxiv, pdf, cication: -1

Se Jin Park, Julian Salazar, Aren Jansen, ..., Yong Man Ro, RJ Skerry-Ryan · (google.github)
TouchTTS: An Embarrassingly Simple TTS Framework that Everyone Can Touch, arXiv, 2412.08237, arxiv, pdf, cication: -1

Xingchen Song, Mengtao Xing, Changwei Ma, ..., Zhendong Peng, Zhiyong Wu
Debatts: Zero-Shot Debating Text-to-Speech Synthesis, arXiv, 2411.06540, arxiv, pdf, cication: -1

Yiqiao Huang, Yuancheng Wang, Jiaqi Li, ..., Shunsi Zhang, Zhizheng Wu
Very Attentive Tacotron: Robust and Unbounded Length Generalization in Autoregressive Transformer-Based Text-to-Speech, arXiv, 2410.22179, arxiv, pdf, cication: -1

Eric Battenberg, RJ Skerry-Ryan, Daisy Stanton, ..., Julian Salazar, David Kao · (sequence-layers - google) · (x)

Voco

Emotion

EmoSphere++: Emotion-Controllable Zero-Shot Text-to-Speech via Emotion-Adaptive Spherical Vector, arXiv, 2411.02625, arxiv, pdf, cication: -1

Deok-Hyeon Cho, Hyung-Seok Oh, Seung-Bin Kim, ..., Seong-Whan Lee

VITS

Efficient

Projects

VibeVoice - microsoft

A Frontier Long Conversational Text-to-Speech Model
marvis-tts - Marvis-Labs
🌟 VibeVoice Technical Report, arXiv, 2508.19205, arxiv, pdf, cication: -1

Zhiliang Peng, Jianwei Yu, Wenhui Wang, ..., Yan Xia, Furu Wei
AudioStory: Generating Long-Form Narrative Audio with Large Language Models, arXiv, 2508.20088, arxiv, pdf, cication: -1

Yuxin Guo, Teng Wang, Yuying Ge, ..., Wei Zou, Ying Shan
tts - inworld-ai
🌟 higgs-audio - boson-ai

Redefining Expressiveness in Audio Generation
MOSS-TTSD - OpenMOSS

Text to Spoken Dialogue Generation
UniTTS - IDEA-Emdoor-Lab

An end-to-end TTS system without decoupling of acoustic and semantic information
🌟 dia - nari-labs
MegaTTS 3: Sparse Alignment Enhanced Latent Diffusion Transformer for Zero-Shot Speech Synthesis, arXiv, 2502.18924, arxiv, pdf, cication: -1

Ziyue Jiang, Yi Ren, Ruiqi Li, ..., Xiang Yin, Zhou Zhao · (sditdemo.github)
csm - SesameAILabs

· (huggingface)
fastrtc - freddyaboulton
Overview of the Amphion Toolkit (v0.2), arXiv, 2501.15442, arxiv, pdf, cication: -1

Jiaqi Li, Xueyao Zhang, Yuancheng Wang, ..., Junan Zhang, Zhizheng Wu
Step-Audio-TTS-3B 🤗
LLaSA: Scaling Train-Time and Test-Time Compute for LLaMA-based Speech Synthesis 🤗
OuteTTS 0.3 1B & 500M 𝕏
tts-generation-webui - rsxdalv
alltalk_tts - erew123
OuteTTS-0.1-350M is a novel text-to-speech synthesis model that leverages pure language modeling without external adapters or complex architectures 🤗

· (outeai) · (𝕏)