# AI use cases for Qwen Audio Agent (QwenAudio/qwen-audio-agent) Repo: [QwenAudio/qwen-audio-agent](https://github.com/QwenAudio/qwen-audio-agent) | Use cases: 4 | Page: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/ qwen-audio-agent is a JavaScript realtime voice runtime that lets people talk with an AI Agent while the Agent runs tool calls and background tasks in the same conversation. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Delegate coding tasks to an Agent by voice: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/voice-delegate-coding-tasks/ - Keep a floating voice assistant on your desktop: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/desktop-floating-voice-orb/ - Talk to an Agent in the browser with live camera frames: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/browser-voice-with-camera/ - Run a voice frontend stack with local speech models: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/local-speech-to-speech-stack/ ## How to delegate coding tasks to an Agent by voice with Qwen Audio Agent Repo: [QwenAudio/qwen-audio-agent](https://github.com/QwenAudio/qwen-audio-agent) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Speak tasks to a backend coding Agent and hear results return to the same conversation. **Who it is for:** A developer who wants to keep talking through a problem while a coding Agent edits files or runs tools in the background. **The problem:** Waiting on a coding Agent to finish a task blocks the conversation, so the developer cannot ask follow-up questions or check progress. **The outcome:** The developer talks to the voice frontend, sends a task to the backend Agent, asks for progress, and gets the result back in the same conversation. ### Steps 1. Install the runtime with `npm install -g qwen-audio-agent`. 2. Create your config and add your API Key with `qwenaudio config`. 3. Set AGENT_PROTOCOL in the config to the backend Agent you want, for example openclaw, or leave it empty for frontend-only mode. 4. Start the Gateway with `qwenaudio` in one terminal. 5. Open a second terminal and start the TUI with `qwenaudio tui`, then speak your task. ### Prompt to paste into your agent ```text Use the repo QwenAudio/qwen-audio-agent to set up a voice session that delegates to a backend coding Agent. Confirm the Gateway starts and the TUI connects, then report the task status returned in the conversation. Verify that the Gateway log shows the task was created before reporting success. ``` ### Caveat The README says the backend Agent must be installed and configured separately, and some backends are marked as not yet fully verified. Risk, credentials: The setup stores a Bailian or other provider API Key in the config file, so keep that file private and do not commit it to a repo. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/voice-delegate-coding-tasks/ ## How to keep a floating voice assistant on your desktop with Qwen Audio Agent Repo: [QwenAudio/qwen-audio-agent](https://github.com/QwenAudio/qwen-audio-agent) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Run a desktop orb that listens for a wake word and handles tasks without opening a terminal. **Who it is for:** A knowledge worker on macOS, Windows, or Linux who wants an always-available voice assistant on the desktop. **The problem:** Switching to a terminal or browser every time you want to ask the Agent something breaks your flow. **The outcome:** You have a persistent floating voice orb with a built-in Gateway that stays idle when unused and responds to the voice wake word. ### Steps 1. Download the installer for your platform from the releases page of the QwenAudio/qwen-audio-agent repo, or build from source. 2. Clone the repo from https://github.com/QwenAudio/qwen-audio-agent and open it in a terminal. (inferred, check the README) 3. Build the desktop app for macOS with `npm run desktop:build:local`, for Windows with `npm run desktop:build:win`, or for Linux with `npm run desktop:build:linux`. 4. Open the installed app and add your API Key in its settings. 5. Say the voice wake word to activate the orb and ask for a task. ### Prompt to paste into your agent ```text Use the repo QwenAudio/qwen-audio-agent to build the desktop app for and launch the floating voice orb. Report the build output path and whether the built-in Gateway starts. Verify the orb responds to the wake word before reporting success. ``` ### Caveat The README describes build steps for each platform, but the Linux build is noted as unsigned and the exact steps for each platform are in the desktop documentation. Risk, credentials: The desktop app needs a provider API Key, so store it only in the app settings and never in shared files. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/desktop-floating-voice-orb/ ## How to talk to an Agent in the browser with live camera frames with Qwen Audio Agent Repo: [QwenAudio/qwen-audio-agent](https://github.com/QwenAudio/qwen-audio-agent) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Use the WebUI to stream microphone audio and bounded camera frames to a visual-capable voice frontend. **Who it is for:** A user who wants a browser-based voice assistant that can see what the camera shows, such as someone describing an object or a workspace. **The problem:** A text or audio-only assistant cannot respond to what is in front of the camera, so the user has to describe everything by hand. **The outcome:** You open the browser UI, speak to the Agent, and the Agent answers using live audio plus bounded camera frames. ### Steps 1. Install the runtime with `npm install -g qwen-audio-agent`. 2. Create your config and add your API Key with `qwenaudio config`. 3. Pick a visual-capable Realtime frontend such as Google Gemini Live or Qwen Omni Realtime and set its key in the config. 4. Start the Gateway with `qwenaudio`. 5. Open the browser UI with `qwenaudio webui` and allow camera and microphone access. ### Prompt to paste into your agent ```text Use the repo QwenAudio/qwen-audio-agent to start the WebUI with a visual-capable Realtime frontend. Report the local WebUI URL and confirm that camera frames are streamed with audio. Verify that the Agent's reply references something visible on camera before reporting success. ``` ### Caveat The README says camera frames are streamed only with a visual-capable Realtime frontend, so choosing a frontend without video input will not give the Agent any camera view. Risk, credentials: The frontend needs a cloud API Key, and camera frames are sent to that cloud service, so use a key and camera setup you trust. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/browser-voice-with-camera/ ## How to run a voice frontend stack with local speech models with Qwen Audio Agent Repo: [QwenAudio/qwen-audio-agent](https://github.com/QwenAudio/qwen-audio-agent) | Difficulty: Advanced | Time: 90 minutes | Basis: From the README Connect the Hugging Face speech-to-speech service so VAD, STT, LLM, and TTS run locally. **Who it is for:** A privacy-minded developer who wants voice conversation without sending audio to a cloud provider. **The problem:** Cloud voice services send audio off the machine, which some users cannot accept for their work. **The outcome:** You run the Hugging Face speech-to-speech service locally, point the runtime at its URL, and have a voice conversation that uses local components. ### Steps 1. Install the runtime with `npm install -g qwen-audio-agent`. 2. Start the Hugging Face Speech-to-Speech service on your machine using its own instructions. 3. Set the voice frontend in the config to Hugging Face Speech-to-Speech and enter the service URL. 4. Start the Gateway with `qwenaudio` and open the TUI with `qwenaudio tui`. 5. Speak a short question and confirm the reply comes from the local service. ### Prompt to paste into your agent ```text Use the repo QwenAudio/qwen-audio-agent to connect the Hugging Face speech-to-speech frontend running at . Report the frontend name shown in the Gateway and whether a spoken question gets a reply. Verify the Gateway log shows the local service URL before reporting success. ``` ### Caveat The README says the frontend is configurable for STT, LLM, and TTS only once the service is running, and the exact service setup is in the linked voice frontend guide, so the first run depends on that separate setup. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/qwenaudio-qwen-audio-agent/local-speech-to-speech-stack/