speakeronnx

speakeronnx is a speaker embedding library built on onnxruntime. It does not need torch at runtime.

The library extracts speaker embeddings from audio, computes cosine similarity between them, and verifies speaker identity. It downloads ONNX models from HuggingFace automatically.

Model collection: OpenVoiceOS/speaker-embeddings-onnx

Install

pip install speakeronnx

Optional high-quality resampling:

pip install speakeronnx soxr

Quick start

from speakeronnx import SpeakerEmbedder, cosine, verify

embedder = SpeakerEmbedder(model="wespeaker-resnet34")

alice1 = embedder.embed("alice_clip1.wav")
alice2 = embedder.embed("alice_clip2.wav")
bob    = embedder.embed("bob_clip1.wav")

print(cosine(alice1, alice2))   # e.g. 0.82  - same speaker
print(cosine(alice1, bob))      # e.g. 0.21  - different speaker

ok, score = verify(alice1, alice2, threshold=0.45)
print(ok, score)  # True 0.82

More examples in examples/.

CLI

speakeronnx list                              # list available models
speakeronnx embed clip.wav                    # extract embedding
speakeronnx verify a.wav b.wav               # same-speaker check (exit 0/1)
speakeronnx verify a.wav b.wav --threshold 0.5
speakeronnx embed clip.wav --model wespeaker-ecapa512

Full CLI reference in docs/cli.md.

Models

The library registers 9 models in MODEL_REGISTRY. Each one downloads on first use:

Full model comparison and selection guide in docs/models.md.

Documentation

Examples

Tests

# Unit tests (mocked, no downloads, no network)
pytest tests/test_unit.py tests/test_audio.py tests/test_frontend.py \
      tests/test_embedder.py tests/test_cli.py tests/test_model_registry.py -v

# End-to-end tests (downloads models + generates TTS audio)
pytest tests/test_e2e.py -v -s

Each registered model has an HF source, license, and embedding dimension:

Run speakeronnx list to print descriptions and metadata for all registered models.

Feature frontend

• fbank80 models use an 80-dim log-Mel filterbank with per-utterance CMN, implemented in pure numpy. See docs/frontend.md.
• raw models (redimnet-b2) pass the raw 16 kHz waveform directly to ONNX. The model has an internal MelSpectrogram.

Audio requirements

• Mono PCM WAV, any bit depth (8/16/24/32-bit int, 32-bit float)
• Any sample rate (the library resamples internally to 16 kHz)
• The library downmixes stereo files to mono
• Minimum length ~1 second
• Recommended enrollment length 5-30 seconds per speaker

Dependencies

• onnxruntime
• numpy
• huggingface_hub
• soxr (optional, for high-quality resampling)

Related projects

• GitHub: TigreGotico/speakeronnx
• PyPI: pip install speakeronnx
• TigreGotico/speechonnxmetrics uses speakeronnx for its speaker-similarity metric.