This is a BF16 conversion of LongCat-AudioDiT-3.5B — a state-of-the-art diffusion-based zero-shot TTS model by Meituan that operates directly in the waveform latent space. Converting from FP32 to BF16 halves the on-disk size and VRAM usage with negligible quality loss, making it the recommended variant for most users.
Original (3.5B FP32)
This (3.5B BF16)
Weight dtype
float32
bfloat16
Activation dtype
float32
bfloat16
File size
~14 GB
~7 GB
VRAM (inference)
~20 GB
~12 GB
Quality
Reference
Virtually identical
Extra dependencies
none
none
Conversion Details
All model weights — DiT transformer backbone, Wav-VAE, and text encoder — are converted from float32 to bfloat16. BF16 preserves the same dynamic range as FP32 (8 exponent bits) while halving memory usage, making it the lossless practical choice for inference on modern GPUs.
No post-training quantization, calibration data, or scale factors are required. The model is a direct dtype cast and is fully compatible with the original audiodit inference code.
Hardware Requirements
GPU: NVIDIA GPU with CUDA support (BF16 supported on Ampere and newer; falls back gracefully on older hardware)
VRAM: ~7 GB
CPU: Supported but slow
Usage — ComfyUI (Recommended)
The easiest way to use this model is with ComfyUI-LongCat-AudioDIT-TTS, which has native support for this BF16 model with zero extra setup.
Installation
Install the ComfyUI node via ComfyUI Manager (search LongCat-AudioDiT) or manually:
dtype: auto or bf16 — matches this model's native dtype
guidance_method: cfg for TTS, apg for voice cloning
steps: 16 (balanced), 32 (higher quality)
keep_model_loaded: True for repeated use
This is the recommended variant for most users — best balance of quality, VRAM usage, and compatibility.
About LongCat-AudioDiT
LongCat-AudioDiT is a non-autoregressive diffusion-based TTS model from Meituan that achieves state-of-the-art zero-shot voice cloning performance on the Seed benchmark. Unlike previous methods relying on mel-spectrograms, it operates directly in the waveform latent space using only a Wav-VAE and a DiT backbone.
The 3.5B variant achieves 0.818 SIM on Seed-ZH and 0.797 SIM on Seed-Hard, surpassing both open-source and closed-source competitors.
