VoxScribe

Offline speech recognition

Built for Unity 2021.3 LTS+C# APIOffline

Convert speech to text fully offline using Whisper-based models. Process microphone input or any AudioClip, buffer or stream into transcripts and translations with partial streaming events for real-time gameplay or chat experiences.

View source Read the docs Contact

SOURCE PREVIEW

A look at the code behind VoxScribe

Assets / Denisitree / VoxScribe / Runtime / VoxScribeRecognizer.cs

C# · UTF-8 · Unity 2021.3+ · 0 lines

Quick start

Use the API in a Unity scene.

Add the VoxScribeRecognizer component to a GameObject.
Set the model path and choose the language or keep it on auto.
Start listening and respond to the OnPartialText and OnFinalText events as results arrive.

Read the docs Contact

What it does

Offline transcription and translation designed for live and local workflows.

Live mic input

Capture speech in real time and stream partial results to the Unity scene or game logic.

AudioClip support

Process recorded or imported AudioClip sources, byte arrays or raw buffers generated elsewhere.

Translation support

Translate spoken content into multiple languages using language auto-detect and model selection.

Streaming callbacks

Send partial transcript updates as they arrive through UnityEvent or C# delegate callbacks.

Model size tuning

Choose model size depending on device power, latency target and desired accuracy.

Unity-ready events

Integrate directly into gameplay systems, editors and tool windows using familiar Unity event patterns.

Planned API

API preview, subject to change.

              Denisitree.VoxScribe
            
using Denisitree.VoxScribe;
using UnityEngine;

public class VoicePrompt : MonoBehaviour
{
    private VoxScribeRecognizer recognizer;

    private void Start()
    {
        recognizer = new VoxScribeRecognizer();
        recognizer.OnPartialResult += text => Debug.Log($"Partial: {text}");
        recognizer.OnFinalResult += text => Debug.Log($"Final: {text}");
        recognizer.Language = "en";
        recognizer.StartMicrophone();
    }
}

Planned platforms & requirements

Runtime plans and expected requirements.

Category	Planned support
Platforms	Windows, macOS, Linux, Android, iOS
Unity	Unity 2021.3 LTS+
Model types	Whisper-based local models, tiny to large
Input sources	Microphone, AudioClip, raw WAV/PCM buffers, byte arrays
Output	Streaming partial transcripts, final transcript, optional translation

Works well with

Pair it with the rest of the Denisitree toolkit.

SonicLoad Vocalis PocketMind

Roadmap

Current focus for VoxScribe.

Core buildWhisper core integration and streaming partial results.

Private alphaLanguage detection, model tuning and UnityEvents feedback.

Public betaProduction stability, sample scenes and performance tuning.

Screenshots