- GPT-Live is now available for all ChatGPT users
- You can talk and listen at the same time for more natural conversations.
- Real-time translations are now also possible
ChatGPT has a brilliant new AI voice model called GPT-Live, which has a number of useful tricks, including the ability to listen and speak at the same time. It is now rolling out to all ChatGPT users, although OpenAI has acknowledged a number of initial bugs.
While free users and users on a paid plan get slightly different models (GPT-Live-1 mini and GPT-Live-1 respectively), the updated model should now appear on all ChatGPT accounts, with the new features described below.
The biggest gift you will get from the upgrade will be the Live label at the top of voice chats on mobile devices and behind the ChatGPT dropdown on the web. Tap or click these labels and you can still revert to the old voice models, for the time being.
There’s another way to check that the GPT-Live voice model has arrived in your account: on mobile, tap the menu button (top left), then the settings gear (Android) or your profile avatar (iOS), and Voice > Model. On the web, click on your profile avatar (bottom left) and then Settings > Voice > Model.
What to try first
The biggest improvement here is the ‘duplex’ functionality, so try that first: you can keep talking even after ChatGPT has started responding to you, and it should continue. Secondly, try to interrupt him mid-stream and he will adapt his response accordingly. We are almost at the level of Spike Jonze’s 2013 film His at this stage.
Third, ask ChatGPT in voice mode to translate something into a foreign language while you say it out loud. You can then speak English sentences and ChatGPT will do a real-time translation without hesitation. It’s not particularly useful for language learning, but it shows off the capabilities of GPT-Live.
Fourth, change the voice and intelligence used; You can do this using the slider icon at the top right of voice chats. The voice options are actually the same as before, but you can choose between Instant, Halfand High like the level of intelligence. Wear Instant for the fastest responses, High to get the best answers and Medium to reach an agreement.
The last thing you can try once you have the update is to ask questions with visual answers. OpenAI has added a bunch of visual cards to voice mode now, so you can get on-screen graphics about sports scores, weather forecasts, and places that can be found on a map, for example.
Early voice errors
We are aware of issues with memory not being as reliable in ChatGPT Voice with GPT-Live. We are actively investigating and will follow up!July 9, 2026
I’ve been testing the GPT-Live voice mode for a few hours now and can report that everything works as advertised. It’s more than ever like talking to a real person, even through the hesitations and variety of speech patterns. I experienced one or two crashes, but they were few and far between.
On Reddit, Atty Eleti from OpenAI is answering questions about GPT-Live. One of the main bugs that users seem to be experiencing is related to ChatGPT memory, which seems to be out of reach of voice mode in some cases; This is an issue that OpenAI is tracking and “actively investigating”, and you can find updates on it here.
Problems are also reported when foreign languages are pronounced with an English accent. Again, this is a problem that has been recognized and should improve over time, according to Eleti.
Overall, though, the implementation seems to be going relatively well and I haven’t seen any problems with memory or accents so far. I’m not sure it’ll make me want to use voice mode more than I already do (which isn’t much), but for regular voice users it’s definitely a big step forward.
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds.
