Speech to Text (STT)

Convert spoken words from an AudioClip into text using powerful AI transcription models.

Ideal for voice commands, user feedback, subtitles, or audio-driven gameplay systems.

---

✅ Basic Usage

AudioClip recording = MicrophoneCapture.GetLastClip();

string result = await recording
    .GENTranscript()
    .SetModel(OpenAIModel.Whisper)
    .SetLanguage(SystemLanguage.Korean)
    .ExecuteAsync();

Debug.Log("Transcript: " + result);

🔊 The AudioClip can be from a microphone, file, or any runtime source.

---

⚙️ Configuration Options

Method	Description
SetLanguage(SystemLanguage)	Optional hint to improve transcription accuracy
SetModel(model)	Choose which STT model to use (Whisper, Gemini STT, etc.)
SetOutputPath(path)	Save transcription to file (optional)

---

🌍 Translation Mode

You can also translate speech into English using GENTranslation():

string english = await recording
    .GENTranslation()
    .SetModel(OpenAIModel.Whisper)
    .ExecuteAsync();

🗣️ This uses the same audio input but produces translated text (into English).

---

📦 Example Result

Audio Input: "안녕하세요, 오늘 날씨 어때요?" Transcript: "안녕하세요, 오늘 날씨 어때요?" Translation: "Hello, how's the weather today?"

---

🧠 Tips

• Works best with clean, mono audio at 16kHz or higher.

• SetLanguage is optional — the model can auto-detect, but accuracy improves with a hint.

• For multilingual games or voice input, pair this with Text Generation for natural response.