# CosyVoice 3: Towards In-the-wild Speech Generation via Scaling-up and Post-training

arXiv: https://arxiv.org/abs/2505.17589 · HTML: https://mlsanity.com/paper/2505.17589/

## Key facts

- Cited by 150 papers in the map (data through 14 Oct 2026).
- Placed in Audio, speech and music, one of 14 directions in Diffusion models.
- Builds on 52 of its 57 references that are in the map.
- First posted on arXiv 23 May 2025.

## Placed in

- Diffusion models / [Audio, speech and music](https://mlsanity.com/diffusion/audio-speech-and-music/)

## Builds on (most cited first, 30 of 52)

- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions](https://mlsanity.com/paper/1712.05884/) (2017)
- [SoundStream: An End-to-End Neural Audio Codec](https://mlsanity.com/paper/2107.03312/) (2021)
- [Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers](https://mlsanity.com/paper/2301.02111/) (2023)
- [Qwen2.5-Omni Technical Report](https://mlsanity.com/paper/2503.20215/) (2025)
- [FastSpeech 2: Fast and High-Quality End-to-End Text to Speech](https://mlsanity.com/paper/2006.04558/) (2020)
- [Tacotron: Towards End-to-End Speech Synthesis](https://mlsanity.com/paper/1703.10135/) (2017)
- [FLEURS: Few-shot Learning Evaluation of Universal Representations of Speech](https://mlsanity.com/paper/2205.12446/) (2022)
- [Finite Scalar Quantization: VQ-VAE Made Simple](https://mlsanity.com/paper/2309.15505/) (2023)
- [Voicebox: Text-Guided Multilingual Universal Speech Generation at Scale](https://mlsanity.com/paper/2306.15687/) (2023)
- [F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching](https://mlsanity.com/paper/2410.06885/) (2024)
- [W2v-BERT: Combining Contrastive Learning and Masked Language Modeling for Self-Supervised Speech Pre-Training](https://mlsanity.com/paper/2108.06209/) (2021)
- [CosyVoice: A Scalable Multilingual Zero-shot Text-to-speech Synthesizer based on Supervised Semantic Tokens](https://mlsanity.com/paper/2407.05407/) (2024)
- [Seed-TTS: A Family of High-Quality Versatile Speech Generation Models](https://mlsanity.com/paper/2406.02430/) (2024)
- [CosyVoice 2: Scalable Streaming Speech Synthesis with Large Language Models](https://mlsanity.com/paper/2412.10117/) (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)
- [Speak, Read and Prompt: High-Fidelity Text-to-Speech with Minimal Supervision](https://mlsanity.com/paper/2302.03540/) (2023)
- [MaskGCT: Zero-Shot Text-to-Speech with Masked Generative Codec Transformer](https://mlsanity.com/paper/2409.00750/) (2024)
- [E2 TTS: Embarrassingly Easy Fully Non-Autoregressive Zero-Shot TTS](https://mlsanity.com/paper/2406.18009/) (2024)
- [Matcha-TTS: A fast TTS architecture with conditional flow matching](https://mlsanity.com/paper/2309.03199/) (2023)
- [VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizers](https://mlsanity.com/paper/2406.05370/) (2024)
- [FunAudioLLM: Voice Understanding and Generation Foundation Models for Natural Interaction Between Humans and LLMs](https://mlsanity.com/paper/2407.04051/) (2024)
- [Spark-TTS: An Efficient LLM-Based Text-to-Speech Model with Single-Stream Decoupled Speech Tokens](https://mlsanity.com/paper/2503.01710/) (2025)
- [EXPRESSO: A Benchmark and Analysis of Discrete Expressive Speech Resynthesis](https://mlsanity.com/paper/2308.05725/) (2023)

## Built on by (most cited first, 30 of 150)

- [Qwen3-Omni Technical Report](https://mlsanity.com/paper/2509.17765/) (2025)
- [Qwen3.5-Omni Technical Report](https://mlsanity.com/paper/2604.15804/) (2026)
- [Qwen3-TTS Technical Report](https://mlsanity.com/paper/2601.15621/) (2026)
- [Streaming Sequence-to-Sequence Learning with Delayed Streams Modeling](https://mlsanity.com/paper/2509.08753/) (2025)
- [OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models](https://mlsanity.com/paper/2604.00688/) (2026)
- [Fun-ASR Technical Report](https://mlsanity.com/paper/2509.12508/) (2025)
- [Fun-Audio-Chat Technical Report](https://mlsanity.com/paper/2512.20156/) (2025)
- [Ming-UniAudio: Speech LLM for Joint Understanding, Generation and Editing with Unified Representation](https://mlsanity.com/paper/2511.05516/) (2025)
- [X-OPD: Cross-Modal On-Policy Distillation for Capability Alignment in Speech LLMs](https://mlsanity.com/paper/2603.24596/) (2026)
- [MOSS-TTS Technical Report](https://mlsanity.com/paper/2603.18090/) (2026)
- [JoyAI-VL-Interaction: Real-Time Vision-Language Interaction Intelligence](https://mlsanity.com/paper/2606.14777/) (2026)
- [Audio-Omni: Extending Multi-modal Understanding to Versatile Audio Generation and Editing](https://mlsanity.com/paper/2604.10708/) (2026)
- [Towards Interactive Intelligence for Digital Humans](https://mlsanity.com/paper/2512.13674/) (2025)
- [Audio2Tool: Speak, Call, Act -- A Dataset for Benchmarking Speech Tool Use](https://mlsanity.com/paper/2604.22821/) (2026)
- [RoboOmni: Proactive Robot Manipulation in Omni-modal Context](https://mlsanity.com/paper/2510.23763/) (2025)
- [VoxSafeBench: Not Just What Is Said, but Who, How, and Where](https://mlsanity.com/paper/2604.14548/) (2026)
- [Step-Audio-EditX Technical Report](https://mlsanity.com/paper/2511.03601/) (2025)
- [Raon-Speech Technical Report](https://mlsanity.com/paper/2605.23912/) (2026)
- [DrVoice: Parallel Speech-Text Voice Conversation Model via Dual-Resolution Speech Representations](https://mlsanity.com/paper/2506.09349/) (2025)
- [ViBES: A Conversational Agent with Behaviorally-Intelligent 3D Virtual Body](https://mlsanity.com/paper/2512.14234/) (2025)
- [UltraEval-Audio: A Unified Framework for Comprehensive Evaluation of Audio Foundation Models](https://mlsanity.com/paper/2601.01373/) (2026)
- [IndexTTS 2.5 Technical Report](https://mlsanity.com/paper/2601.03888/) (2026)
- [NV-Bench: Benchmark of Nonverbal Vocalization Synthesis for Expressive Text-to-Speech Generation](https://mlsanity.com/paper/2603.15352/) (2026)
- [MMA-ASIA: A Multilingual and Multimodal Alignment Framework for Culturally-Grounded Evaluation](https://mlsanity.com/paper/2510.08608/) (2025)
- [ReStyle-TTS: Relative and Continuous Style Control for Zero-Shot Speech Synthesis](https://mlsanity.com/paper/2601.03632/) (2026)
- [ID-LoRA: Identity-Driven Audio-Video Personalization with In-Context LoRA](https://mlsanity.com/paper/2603.10256/) (2026)
- [FlowTTS-GRPO: Online Reinforcement Learning with Multi-Objective Reward Optimization for Flow-Matching Based Text-to-Speech](https://mlsanity.com/paper/2606.23190/) (2026)
- [Direct Preference Optimization for Speech Autoregressive Diffusion Models](https://mlsanity.com/paper/2509.18928/) (2025)
- [Selective Classifier-free Guidance for Zero-shot Text-to-speech](https://mlsanity.com/paper/2509.19668/) (2025)
- [Bagpiper: Solving Open-Ended Audio Tasks via Rich Captions](https://mlsanity.com/paper/2602.05220/) (2026)

## Abstract (arXiv)

In our prior works, we introduced a scalable streaming speech synthesis model, CosyVoice 2, which integrates a large language model (LLM) and a chunk-aware flow matching (FM) model, and achieves low-latency bi-streaming speech synthesis and human-parity quality. Despite these advancements, CosyVoice 2 exhibits limitations in language coverage, domain diversity, data volume, text formats, and post-training techniques. In this paper, we present CosyVoice 3, an improved model designed for zero-shot multilingual speech synthesis in the wild, surpassing its predecessor in content consistency, speaker similarity, and prosody naturalness. Key features of CosyVoice 3 include: 1) A novel speech tokenizer to improve prosody naturalness, developed via supervised multi-task training, including automatic speech recognition, speech emotion recognition, language identification, audio event detection, and speaker analysis. 2) A new differentiable reward model for post-training applicable not only to CosyVoice 3 but also to other LLM-based speech synthesis models. 3) Dataset Size Scaling: Training data is expanded from ten thousand hours to one million hours, encompassing 9 languages and 18 Chinese dialects across various domains and text formats. 4) Model Size Scaling: Model parameters are increased from 0.5 billion to 1.5 billion, resulting in enhanced performance on our multilingual benchmark due to the larger model capacity. These advancements contribute significantly to the progress of speech synthesis in the wild. We encourage readers to listen to the demo at https://funaudiollm.github.io/cosyvoice3.