---
license: apache-2.0
language:
  - en
pipeline_tag: text-to-speech
base_model: owensong/Inflect-Micro-v2
model_name: Inflect-Micro-v2-ONNX
tags:
  - onnx
  - onnxruntime
  - text-to-speech
  - speech-synthesis
  - local-tts
  - cpu
  - cuda
  - directml
  - edge-ai
  - small-model
  - vits
  - 24khz
inference: false
---

# Inflect-Micro-v2 ONNX

Official, verified FP32 ONNX Runtime export of
[Inflect-Micro-v2](https://huggingface.co/owensong/Inflect-Micro-v2).
This repository is a **format export only**: no training, pruning, or
quantization was applied.

- **Complete neural weights:** approximately 37.75 MB
- **Output:** 24 kHz mono waveform
- **Providers:** CPU, CUDA, or DirectML through ONNX Runtime
- **Frontend:** Python and eSpeak-ng
- **License:** Apache 2.0

The learned model is split into:

| Graph | Purpose | Size |
| --- | --- | ---: |
| [`onnx/duration.onnx`](onnx/duration.onnx) | Tokens to aligned acoustic distribution | 7.32 MB |
| [`onnx/decode.onnx`](onnx/decode.onnx) | Acoustic distribution and seeded noise to waveform | 30.43 MB |

## Quick start

```bash
python -m pip install -r onnx/requirements.txt
python onnx/inference_onnx.py \
  --text "A small voice can still have something meaningful to say." \
  --output sample.wav \
  --provider cpu
```

See the complete [ONNX guide](onnx/README.md) for the Python API, provider
selection, long-text behavior, parity measurements, checksums, and
reproducible export instructions.

## Verification

The graphs pass ONNX checker validation, dynamic-length inference,
short- and long-form synthesis, seed determinism, and direct numerical parity
testing against the canonical PyTorch checkpoint. See
[`onnx/parity_report.json`](onnx/parity_report.json) and
[`onnx/SOURCE.json`](onnx/SOURCE.json).

The canonical architecture, evaluation, audio samples, PyTorch checkpoint, and
project documentation remain at
[owensong/Inflect-Micro-v2](https://huggingface.co/owensong/Inflect-Micro-v2).