audiopipe

Fast speech-to-text in Rust. Supports ONNX Runtime and GGML backends — runs on Metal (macOS GPU), Vulkan (Windows/Linux GPU), CUDA, CoreML, DirectML, or CPU. No Python, no ffmpeg.

use audiopipe::{Model, TranscribeOptions};

let mut model = Model::from_pretrained("qwen3-asr-0.6b-ggml")?;
let result = model.transcribe(&audio_f32, 16000, TranscribeOptions::default())?;
println!("{}", result.text);

Models

Model	Backend	Params	Languages	Notes
qwen3-asr-0.6b-ggml	GGML	0.6B	6000+	Recommended. Metal/Vulkan/CUDA GPU support
qwen3-asr-0.6b	ONNX	0.6B	6000+	CoreML/DirectML/CUDA via ONNX Runtime
parakeet-tdt-0.6b-v2	ONNX	0.6B	English	NVIDIA Parakeet TDT
parakeet-tdt-0.6b-v3	ONNX	0.6B	25 languages	NVIDIA Parakeet TDT
whisper-*	whisper.cpp	varies	99 languages	e.g. whisper-large-v3-turbo

Usage

Qwen3-ASR GGML (recommended)

Best balance of speed, accuracy, and language coverage. Uses GGML backend with native GPU support.

[dependencies]
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr-ggml"] }

GPU acceleration:

# macOS — Metal GPU
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr-ggml", "metal"] }

# Windows/Linux — Vulkan GPU
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr-ggml", "vulkan-ggml"] }

# NVIDIA — CUDA GPU
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr-ggml", "cuda-ggml"] }

Qwen3-ASR ONNX

[dependencies]
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr"] }

# With CoreML (macOS)
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr", "coreml"] }

# With DirectML (Windows)
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["qwen3-asr", "directml"] }

Parakeet

[dependencies]
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["parakeet"] }

Whisper

[dependencies]
audiopipe = { git = "https://github.com/screenpipe/audiopipe.git", features = ["whisper"] }

Feature flags

Feature	Description
qwen3-asr-ggml	Qwen3-ASR via GGML (recommended)
qwen3-asr	Qwen3-ASR via ONNX Runtime
parakeet	NVIDIA Parakeet TDT via ONNX Runtime
whisper	Whisper via whisper.cpp
metal	Metal GPU for GGML models (macOS)
vulkan-ggml	Vulkan GPU for GGML models (Windows/Linux)
cuda-ggml	CUDA GPU for GGML models (NVIDIA)
coreml	CoreML for ONNX models (macOS)
directml	DirectML for ONNX models (Windows)
cuda	CUDA for ONNX models (NVIDIA)

Examples

# Transcribe with GGML Qwen3-ASR
cargo run --example test_ggml --features qwen3-asr-ggml -- audio.wav

# Transcribe with Parakeet
cargo run --example transcribe --features parakeet -- audio.wav

# Benchmark
cargo run --example benchmark --features qwen3-asr-ggml --release

Used by

• screenpipe — AI screen + audio memory for your desktop

License

MIT