> ## Documentation Index > Fetch the complete documentation index at: https://docs.gladecore.com/llms.txt > Use this file to discover all available pages before exploring further. # Step 4: Speech to Text (STT) > Follow the instructions below to set up and use Speech to Text (STT) capabilities. ## General Setup 1. Go to ChatUICanvas in the DemoChat Scene. Find STTManager Script. *Note: if you can't find it, add \[STTManager] script, assign it to \[NPCChatUI].* Unity STT Manager Script Unity Chat UI Canvas STT Manager Unity Voice Input STT Auto Send Recognized Text 2. **Auto Send Recognized Text:** Checked by default. The recognized STT text is sent to the chat automatically. 3. **Disable Text Input While Recording**: Checked by default. The user cannot type in the chat while recording. ## STT Manager Settings 1. Check Enable Voice Input to enable STT Unity STT Settings Enable Voice Input 2. Model Path * Download whisper models from [https://huggingface.co/ggerganov/whisper.cpp/tree/main](https://huggingface.co/ggerganov/whisper.cpp/tree/main) * Copy the model file to `/Assets/StreamingAssets/Whisper/` * Select the model for STT Unity Whisper Model Settings Model Path Unity Whisper Model Path Set GGML **Is Models in Streaming Assets**: checked by default so the packaged game includes the Whisper models. 3. Select the Language Translate To English is checked by default. *Note: If set to Auto, Whisper may mis-detect the spoken language and produce garbled output.* Unity STT Language Select English 4. Select Voice Input Key Press and hold the voice input key to record. Release to end recording. Unity STT Select Voice Input Key 5. Microphone Settings Unity STT Microphone Settings 6. Debug Unity STT Debug Log STT