---
library_name: onnx
tags:
  - moonshine
  - speech-to-text
  - automatic-speech-recognition
  - audio
  - streaming
  - onnx
  - inference4j
license: mit
pipeline_tag: automatic-speech-recognition
---

# Moonshine Streaming (Tiny) — ONNX Runtime (.ort)

Moonshine v2 streaming tiny speech recognition model for English, quantized to ONNX Runtime (.ort) format for efficient on-device transcription.

Mirrored for use with [inference4j](https://github.com/inference4j/inference4j), an inference-only AI library for Java.

## Original Source

- **Repository:** [Moonshine AI (moonshine-ai/moonshine)](https://huggingface.co/moonshine-ai/moonshine-voice-assets)
- **License:** mit

## Usage with inference4j

```java
try (var recognizer = MoonshineRecognizer.builder()
        .modelId("inference4j/moonshine-streaming-tiny-en")
        .build()) {
    RecognitionResult result = recognizer.recognize(audioStream);
    System.out.println(result.text());
}
```

## Model Details

| Property | Value |
|----------|-------|
| Architecture | Moonshine v2 Streaming (Tiny) |
| Release | quantized_26_08_21 |
| Task | Streaming automatic speech recognition |
| Precision | Quantized (.ort) |
| Language | English (en) |
| Limits | At most 30 s per call (hard graph limit: 4096 frames / ~82 s) |

## License

This model is licensed under the [MIT License](https://opensource.org/licenses/MIT). Original models by [Moonshine AI](https://github.com/moonshine-ai/moonshine).

## Limits

- **Language:** English only.
- **Audio duration:** At most 30 s per call; use `SegmentedSpeechRecognizer` for longer audio.
- **Graph frame limit:** Hard limit inside the graphs is 4096 frames (about 82 s).