AuK

AuK is an open-source foundational model for speech generation and editing across multiple tasks.

AuK is a 1.5B foundation model for speech generation and editing trained on millions of hours of diverse audio data. It supports zero-shot and instruction-based text-to-speech, content and acoustic editing, paralinguistic editing, speech enhancement, and source separation through a unified natural-language instruction interface. The product features two variants including the base model for high-quality generation and AuK-Flash, a distilled model for fast four-step inference. It provides a prompt enhancer, command-line inference tools, an interactive Gradio demo, and ComfyUI integration. It runs on consumer GPUs with local inference support.

AuK at a glance

Best for: Software engineer, Social media creator
Runs on: Web, CLI, self-hosted, open source
Pricing: Free
Released: August 2026
Category: AI audio tool
Website: github.com/Tencent-Hunyuan/AuK

Key features

Perform zero-shot and instruct-based text-to-speech
Edit speech content by replacing, inserting, or removing text
Adjust pitch, speed, and volume of audio recordings
Modify paralinguistic features like emotion, timbre, and accent
Convert between normal speech and whisper
Denoise, dereverberate, and restore natural speech
Perform speech, music, and target speaker separation

Use it for

Generate custom speech from text and voice descriptions
Edit spoken content or lyrics in existing audio
Enhance audio quality by removing noise and reverberation
Extract singing voices or specific speakers from a mix

Alternatives to AuK

Vosk — It enables offline, low-latency speech recognition on small devices.
AssemblyAI — Industry-leading speech models deliver high accuracy and low latency for voice applications.
WeNet — Achieves state-of-the-art results on public speech datasets.
MOSS-TTS — Delivers zero-shot voice cloning and fine-grained pronunciation control.
VoxCPM — Provides tokenizer-free multilingual speech with controllable voice creativity.
Kits — Ethically licensed voice models and royalty-free music production tools.
Proteus — Adds audio capability to DeepSeek Harness without code changes.
Speech-to-Speech — Fully modular, swappable components with local or hosted execution options.
audio.cpp — Runs modern audio models locally with high-performance, native C++ execution.
Fish Audio — Provides studio-grade voice cloning and text-to-speech with emotion control.
FluidAudio — Runs local audio AI models on Apple Neural Engine.
FireRedAudio — A unified model handling understanding and generation with decoupled continuous representations.

All AuK alternatives