LocalClaw/TTS/LongCat-AudioDiT 3.5B
Local TTS model

LongCat-AudioDiT 3.5B

Catalogue summary: MIT-licensed diffusion TTS model that operates in waveform latent space. The official release ships a Hugging Face-compatible 3.5B safetensors checkpoint, code, CLI and Python inference for high-quality Chinese and English zero-shot voice cloning.

Repository editorial metadata; verify comparative claims in the linked upstream material.

GPU recommendedtext-to-speech generation2 languagesMIT
Choose an app

Start with Recommended. No terminal commands are shown.

Compare speech models
Open model filesRecommended verified starting pointOpen in UnslothLaunches Unsloth Desktop on this modelOpen official projectApplication source and guided setupOpen on Hugging FaceFiles, licence and model cardCheck my machineOptional: verify the fit with your saved hardware

Desktop app links require the app to be installed. If nothing opens, LocalClaw will show app-download and model-file fallbacks.

Catalogue quality
9.5/10
Catalogue speed
7.1/10
Model size
3.5B params; single safetensors checkpoint
Voices
Zero-shot reference-audio cloning

Can LongCat-AudioDiT 3.5B run locally?

LongCat-AudioDiT 3.5B can generate speech locally for private voice workflows. Use the verified setup options on this page; no terminal command is required to choose the right path.

MIT license. Still verify upstream usage notes before shipping.

Upstream source
cloningmultilingual

Audio profile

Cat. quality
9.5
Cat. speed
7.1
Audio
8.7

Best fit

LongCat-AudioDiT 3.5B is best for local voice cloning and expressive speech generation.

Hardware: gpu

Model details

Type
Local TTS model
Family
longcat
Latency
medium
Formats
transformerspytorchsafetensors
Languages
zh, en
Context
3.5B MIT weights, 24 kHz output, official CLI and Python API; README benchmarks Chinese and English Seed tasks and requires a CUDA-style PyTorch path

Install locally

01
Check runtimeConfirm the backend supports transformers, pytorch, safetensors on your machine.
02
Open recommended setupUse the app and model links above. LocalClaw does not expose a terminal command.
03
Test locallyRun a short private audio prompt before moving into production workflows.

Good for

  • text-to-speech generation
  • GPU recommended local workflows
  • cloning, multilingual

Watch before shipping

  • Validate pronunciation, latency and artifacts with your own voice samples.
  • Review the upstream license and acceptable-use notes.
  • Benchmark on your target CPU, Apple Silicon or GPU setup.

Related TTS and speech models

Mistral AI Voxtral 4B TTS 2603 Local TTS model · Q 9.5 · Speed 9.1 Phạm Nguyễn Ngọc Bảo VieNeu-TTS v3 Turbo Local TTS model · Q 9.2 · Speed 9.4 Samuel Vitorino / HALO Research Sopro V2 Turbo Local TTS model · Q 9.2 · Speed 9.4 Zyphra ZONOS2 Local TTS model · Q 9.6 · Speed 8.5 Audio8 Audio8 TTS Preview 0.6B Local TTS model · Q 9.3 · Speed 8.9 Bilibili IndexTTS 2.5 Local TTS model · Q 9.5 · Speed 8.4 Alibaba FunAudioLLM CosyVoice 2 Local TTS model · Q 9.3 · Speed 8.8 Boson AI Higgs TTS 3 4B Local TTS model · Q 9.7 · Speed 8
CompareBrowse all TTS models Local AIBrowse LLM models macOS appGet LocalClaw