audio.cpp

audio.cpp is an audio inference framework that runs modern models locally using a high-performance C++ engine built on ggml without Python dependencies.

audio.cpp is a native C++ inference framework built on top of ggml for running audio models. It eliminates Python dependencies by providing a shared native runtime for tasks including text-to-speech, speech-to-text, voice cloning, music generation, and source separation. The framework supports over 100 models and various GGUF variants, providing performance optimizations for NVIDIA, AMD, and Apple Silicon hardware. Users interact with the software through a shared CLI or server, which facilitates consistent execution across different model families. By utilizing GGUF loading and optimized C++ backends, the project achieves faster inference speeds compared to standard Python-based implementations while reducing memory usage.

audio.cpp at a glance

Best for: Software engineer
Runs on: Windows, Linux, macOS, CLI, self-hosted
Pricing: Free
Released: June 2026
Category: AI audio tool
Website: github.com/0xShug0/audio.cpp

Key features

Native C++ inference runtime
GGUF model format support
Hardware acceleration for NVIDIA, AMD, and Apple Silicon
Side-by-side model comparison via Arena UI
Batch-style offline inference
Experimental JSON pipeline support
Built-in audio resampling and denoising utilities

Use it for

Text-to-speech synthesis
Speech-to-text transcription
Voice cloning and conversion
Music and song generation
Audio source separation and diarization

Alternatives to audio.cpp

transcribe.cpp — Runs 16 model families and 60+ variants with GPU and CPU acceleration.
whisper.cpp — It provides high-performance, offline speech recognition on diverse hardware.
FluidAudio — Runs local audio AI models on Apple Neural Engine.
Samples — Generates local audio tracks and sound effects with Stable Audio 3.
Dlgo — Run LLM inference quickly in pure Go without Python or external runtimes.
AuK — It handles zero-shot TTS, editing, and enhancement via natural language.
MLX Audio — It enables fast audio model inference directly on Apple M-series chips.
AssemblyAI — Industry-leading speech models deliver high accuracy and low latency for voice applications.
stable-diffusion.cpp — Runs diffusion models efficiently in pure C and C++ with no external dependencies.
WeNet — Achieves state-of-the-art results on public speech datasets.
sherpa-onnx — Runs complex audio AI tasks offline on local devices and embedded hardware.
kokoro-onnx — Runs local TTS inference with efficient ONNX runtime performance.

All audio.cpp alternatives