- The advanced voice mode of ChatGPT It becomes integrated into the same chat window as the text.
- Users can speak, view real-time transcripts, and receive images, maps, or other visuals all in one interface.
- The option to activate "Separate Mode" remains to preserve the classic virtual assistant-style audio-only experience.
- The update is being rolled out on the web and in the apps mobiles of iOS and Android for all users, with extra features for paid accounts.
The assistant OpenAI It takes an important step in the way we relate to the Artificial Intelligence. Hereinafter, ChatGPT's voice mode it ceases to be a separate screen and it now coexists directly with the text chat, simplifying daily use and eliminating many unnecessary window switches.
With this update, anyone can talk to ChatGPT, see the transcript of the conversation and receive maps, images, or other visual content without leaving the same chat thread. The idea is to bring the experience closer to a natural conversation, where voice and screen work together instead of being separate.
Voice mode within chat: what exactly changes
Until now, those who wanted to use voice in ChatGPT had to skip to a dedicated audio interfacedominated by the classic blue orb or a full-screen mode distinct from the usual chat. This generated some friction, especially if the user wanted to review previous messages or consult visual information while speaking.
With the new version, Advanced voice mode is activated directly from the typing bar.By tapping the sound wave icon to the right of the text box. There's no abrupt change of environment: the same conversation thread and the complete history remain visible.
As soon as that icon is pressed, ChatGPT starts listening and It displays a live transcript of what is being said on the screen.from both the user and the assistant. The result is a hybrid experience where you can follow the conversation by voice without losing control of what appears in the chat.
The integration also allows the assistant to interact during the dialogue. Add real-time visuals such as maps, related images, web page snippets, or other resources. All of this is presented embedded in the same thread, without having to leave voice mode or open additional windows.
One practical detail is that You can alternate between writing and speaking continuouslyEven if voice mode is active, if the user prefers to type part of the query, the system will accept it and respond by voice, maintaining the continuity of the conversation.
A more natural and faster experience: latency, emotions and GPT-5.1
Voice and text integration doesn't happen on its own. OpenAI has introduced Technical adjustments to make voice interaction smootherwith response times that approximate the pace of a conversation between two people. The company reports responses in the region of 200 milliseconds, which significantly reduces the feeling of waiting.
At the same time, the assistant incorporates improvements in intonation and expressiveness of voicesThe goal is to make them sound less robotic and more like an everyday conversation. The idea is for the user to perceive a more personal tone, capable of conveying subtle nuances and emotions while still remaining an automated tool.
On a technical level, these new features rely on integration with newer models, such as GPT-5.1which allow for more precise adjustment of the pitch, speed, and the way in which the IA It responds via audio. Although these advances don't transform the assistant into a human interlocutor, they do reduce some of the distance typically associated with synthetic voices.
This approach fits with the industry trend towards richer multimodal interactions, in which text, voice, and images are combined into a single stream. Compared to rival solutions such as Gemini Live from GoogleOpenAI's approach is to integrate everything into the same interface, instead of forcing users to jump from one context to another.
For the end user, the practical consequence is that can hold a hands-free conversation much more continuouswhile also obtaining visual support when the consultation requires it, whether to orient oneself with a map, review a graph or follow a diagram on screen.
How to activate it, on which devices, and differences between free and paid users
The new voice experience is gradually rolling it out on both the web and mobile apps from ChatGPT for iOS and AndroidIn most cases, simply updating the app from the corresponding store or refreshing the web version will make the change available.
Once the latest version is installed, access is simple: Simply tap the voice wave icon next to the text box from the chat. From that moment on, the application listens to the user and displays the transcript and responses in the window, without changing screens.
For those using the free version of the service, online voice mode is available. Available at no extra cost, although with limitations in There of use if you don't have a paid subscription. However, plans like ChatGPT Plus, Pro, or Teams offer more chat minutes and access to an advanced voice mode with more elaborate voices and enhanced audio capabilities.
They do, in fact, exist two distinct voice experiences: a standard one, accessible to any user, based on more conventional recognition and synthesis technologies; and an advanced one, which takes advantage of the capabilities of more powerful models to offer more expressive responses and a more polished interaction in real time.
In Spain and the rest of Europe, the update follows the same pattern as in other markets: It is being activated gradually on mobile devices and on the webTherefore, not all users receive it on the same day. Even so, OpenAI indicates that the rollout is designed to reach all accounts, without regional restrictions, beyond the difference between free and paid plans.
More user control: "Separate mode" and voice settings
Text and voice integration is the default approach, but OpenAI has not eliminated the classic audio-only experienceFor those who prefer a more immersive interaction, without seeing the chat or the transcript, there is still the option to use the so-called "Separate Mode".
This mode can be enabled from the ChatGPT settings menu, in the Voice Mode sectionWhen activated, the application reverts to the previous design, in which the user enters an environment dedicated exclusively to audio conversation, similar to talking to a traditional digital assistant.
Switching between integrated interface and separate mode There is no limit to the number of activations.The user can try one, return to the other, and adjust the settings as many times as they like. This flexibility aims to cater both to those who value having their chat history always visible and to those who are more comfortable with a clean, voice-focused screen.
In addition to choosing the type of interface, it is possible to access the settings customize some aspects of the voicesuch as the selection between different available voices. In the advanced mode, these voices have been designed to sound more natural and with a slightly richer intonation, while still retaining their function as an assistance tool.
The fact that the company is maintaining both options reflects a certain degree of caution: Not all users immediately accept design changesAnd the transition to a single interface may generate resistance among those who had already become accustomed to the previous flow. Therefore, the update offers new features without closing the door to previous habits.
Impact on productivity, startups and use cases in Europe
Unifying voice and text in a single window not only improves the convenience for home users; it also opens up new possibilities for startups and teams working with automationBeing able to combine dictation, spoken responses, and visual content in a single interface simplifies the creation of assistants and conversational tools.
In the European context, this integration can be especially useful in hybrid and remote work environmentswhere the ability to make quick voice queries while reviewing documents, maps, or dashboards on screen is increasingly valued. Sectors such as customer service, online education, or technical support can benefit from this multimodal approach.
For founders and technical teams, having access to A single environment for text and voice facilitates proof of concept and the development of products that integrate voice input without the need to design separate interfaces. Even with no-code tools, it's easier to experiment with assistants that combine dictation, spoken responses, and visual elements within the same workflow.
Furthermore, the presence of a voice mode accessible from the web version and mobile apps lowers accessibility barriers For people who prefer not to type on the keyboard or who have visual difficulties, being able to hear the answers without giving up visual information when they need it.
The move fits into an AI industry that, both in Spain and in the rest of the continent, is experiencing a moment of expansion in use and investmentThe major platforms, including OpenAI, are competing to offer more complete and easier-to-adopt experiences, aware that small improvements in usability can make all the difference in mass adoption.
With this change, ChatGPT takes another step towards a A truly multimodal interaction, where speaking, reading, and viewing content happen in the same placeThe option to choose between an integrated interface or a separate mode, combined with improvements in speed and naturalness of the voices, puts the assistant in a more comfortable position for daily use, both by individual users and organizations looking to introduce voice into their workflows without additional complications.
Passionate writer about the world of bytes and technology in general. I love sharing my knowledge through writing, and that's what I'll do on this blog, show you all the most interesting things about gadgets, software, hardware, tech trends, and more. My goal is to help you navigate the digital world in a simple and entertaining way.