ChatGPT Voice has learned a useful new trick. When you're using the new Live voice model it can now use plugins and connected apps while you are talking to it, bringing services and brands into a spoken conversation.
I spent some time trying the combination with everyday tasks. Instead of asking ChatGPT about dinner and then opening a grocery app myself, for example, I can keep going by voice.
The change sounds relatively small until you use it because plugin support means some of those conversations can now lead somewhere beyond ChatGPT itself.
To use plugins with Voice you need to select the Live voice model. To make sure you're using Live go to in and make sure it is selected.
There is a little setup required first. Plugins have to be available on your ChatGPT account, and some need to be connected or require you to sign into the service behind them. I would do that before starting Voice mode, because interrupting a supposedly hands-free conversation to deal with logins rather ruins the effect. Go to then to do this.
Asking for dinner and music
Once I had everything ready, select Voice mode by tapping the Voice button at the right of the prompt window.
I started with asking ChatGPT Voice to use the plugin, as it's relatively easy to converse about the food I want. I told ChatGPT that I needed something easy for dinner and described some ingredients I already had. ChatGPT Voice worked well here as I could keep talking and interrupt myself as I looked through the fridge. ChatGPT suggested a recipe and the ingredients I would need and used Instacart to put together an order for me.
Normally, this process involves bouncing between a recipe, the refrigerator, and a grocery app. With Voice and , more of that happened in one conversation. I still had to look at the screen when an action needed my approval, but I wasn't manually carrying the results from ChatGPT into another app.
Voice doesn't let you verbally wave through every consequential action a plugin might take. When something requires confirmation, ChatGPT puts it on the screen for you to approve. Besides that last element, it was all easy to do over voice.
And it does have its limits with which plugins it can use and how. isn't available, for instance, but is. That said, while Apple Music's plugin can find music and make playlists, it can't leverage your own library or recent listening. It was nice to make a playlist through conversation, but that aspect somewhat limited its appeal to me.
worked better in both narrowing down a search from generally looking for something to do to a specific event and time. I told ChatGPT I wanted something to do and gave it a rough idea of the kind of evening I had in mind. From there, I could narrow things down based on what sounded appealing.
That may be one of the more useful ways to approach plugins in Voice. You don't necessarily need to arrive knowing precisely what you want. It actually fits the experience better to figure it out through conversation as that's one of the few things Voice does better than looking at filters on a website.
Conversational extensions by AI
Plugins don't suddenly make the screen obsolete, and I wouldn't want them to. I want to see what I'm doing before hitting the execute button, especially if it's a purchase. Still, using ChatGPT Voice does change everything leading up to that point.
There are limitations. You only get the plugins available to your account, and availability can depend on your plan. I would also still be selective about what plugins I connect, whether by voice or otherwise. But for what it does, being able to converse by voice is genuinely appealing.
ChatGPT Voice has always been good at talking. Giving it plugins means the conversation doesn't have to end just when it becomes useful. ChatGPT Voice may actually be the easiest way to do some things beyond just engaging with ChatGPT without typing.
Eric Hal Schwartz is a freelance writer for TechRadar with more than 15 years of experience covering the intersection of the world and technology. For the last five years, he served as head writer for Voicebot.ai and was on the leading edge of reporting on generative AI and large language models. He's since become an expert on the products of generative AI models, such as OpenAI’s ChatGPT, Anthropic’s Claude, Google Gemini, and every other synthetic media tool. His experience runs the gamut of media, including print, digital, broadcast, and live events. Now, he's continuing to tell the stories people want and need to hear about the rapidly evolving AI space and its impact on their lives. Eric is based in New York City.