LongCat AudioDiT 3.5B
LongCat-AudioDiT-3.5B is Meituan LongCat’s diffusion-based text-to-speech model built directly in waveform latent space rather than mel-spectrogram space. It is designed for high-fidelity speech generation and zero-shot voice cloning, supports Chinese and English, and is positioned as a top-performing open model on the Seed benchmark for speaker similarity and intelligibility.
- Type
- Audio
- Weights
- Open
- Downloads
- 1.2k
- Likes
- 77
- License
- mit
Model hub
meituan-longcat/LongCat-AudioDiT-3.5B
- Repo
- meituan-longcat/LongCat-AudioDiT-3.5B
- Access
- Public
Technical tags and languages
More from Meituan
11
-
1.6k DL
LongCat 2.0
Text
LongCat-2.0 is an open-source MoE language model with 1.6 trillion total parameters (~48B activated per token…
-
24k DL
LongCat Flash Chat
Text
Chat-oriented LongCat-Flash variant, a 560B MoE language model with around 27B parameters active per token, t…
-
186 DL
LongCat Flash Omni
Multimodal
560B parameter omni-modal MoE model (about 27B active) for real time audio-visual interaction, built on LongC…
-
216 DL
LongCat Flash Thinking
Text
Large reasoning model built on a 560B MoE backbone, activating about 18.6 to 31.3B parameters for advanced ch…
-
14k DL
LongCat Image
Image
6B parameter bilingual (Chinese-English) text-to-image foundation model focused on photorealism, strong Chine…
-
265 DL
LongCat Next
Multimodal
LongCat Next is a multimodal LongCat model focused on compact yet capable visual and speech understanding. Th…
-
2k DL
LongCat Video Avatar 1.5
Multimodal
LongCat-Video-Avatar-1.5 is Meituan LongCat’s open-source audio-driven avatar video model for single-characte…
-
1.7k DL
LongCat-Video
Video
LongCat-Video is a long-horizon text and image to video model that keeps identity, style, and motion consiste…
About Meituan
1
Chinese local services platform connecting consumers with food delivery, restaurant booking, travel, retail, grocery, and on-demand neighborhood services.