CosyVoice 2.0-0.5B

Multilingual Support

• Supported Languages: Chinese, English, Japanese, Korean, Chinese dialects (Cantonese, Sichuanese, Shanghainese, Tianjinese, Wuhanese, etc.)
• Cross-lingual & Mixed-lingual: Supports zero-shot voice cloning for cross-language and code-switching scenarios.

Ultra-Low Latency

• Bidirectional Streaming Support: CosyVoice 2.0 integrates offline and streaming modeling technologies.
• Rapid First Packet Synthesis: Achieves latency as low as 150ms while maintaining high-quality audio output.

High Accuracy

• Improved Pronunciation: Reduces pronunciation errors by 30% to 50% compared to CosyVoice 1.0.
• Benchmark Achievements: Attains the lowest character error rate on the hard test set of the Seed-TTS evaluation set.

Strong Stability

• Timbre Consistency: Ensures reliable voice consistency for zero-shot and cross-language speech synthesis.
• Cross-language Synthesis: Shows significant improvements compared to version 1.0.

Natural Experience

• Enhanced Prosody and Sound Quality: Improved alignment of synthesized audio, raising MOS evaluation scores from 5.4 to 5.53.
• Emotional and Dialectal Flexibility: Now supports more granular emotional controls and accent adjustments.

多语言支持

• 支持的语言：中文、英语、日语、韩语、中国方言（粤语、四川话、上海话、天津话、武汉话等）
• 跨语言与混合语言：支持跨语言和代码切换场景下的零样本声音克隆。

超低延迟

• 双向流式支持：CosyVoice 2.0 集成了离线和流式建模技术。
• 快速首包合成：在保持高质量音频输出的同时，实现低至 150ms 的延迟。

高精度

• 改进的发音：与 CosyVoice 1.0 相比，发音错误减少了 30% 到 50%。
• 基准测试成就：在 Seed-TTS 评估集的困难测试集上获得最低字符错误率。

强大的稳定性

• 音色一致性：确保零样本和跨语言语音合成的可靠声音一致性。
• 跨语言合成：与 1.0 版本相比有显著改进。

自然体验

• 增强的韵律和音质：改进了合成音频的对齐，将 MOS 评估分数从 5.4 提高到 5.53。
• 情感和方言灵活性：现在支持更细粒度的情感控制和口音调整。 来一段中英双语的 英语在前