ChatGPT Voice Gets a Big Upgrade: Now Works Inside Your Chat

ChatGPT Voice just received a major upgrade, and it changes how users talk to the AI. OpenAI now lets you use the feature directly inside an ongoing chat instead of opening the older floating interface. Moreover, the new layout feels faster and more natural for real conversations.

Live Transcript Pops Up As You Speak

Now, ChatGPT shows a live transcript of your voice in real time. This update makes the experience smoother and more transparent. Additionally, users can follow each part of the conversation without switching screens.

Visuals Join the Conversation

In the new mode, ChatGPT Voice also displays visuals linked to the chat. For instance, you might see maps, reference photos, or helpful images based on what you ask. In OpenAI’s demo, the AI pulled up bakery locations and pastry photos during a simple voice query. This move finally combines voice and visuals in one seamless flow.

Switch Back If You Prefer the Old Style

OpenAI also added a simple toggle for people who want the original orb-style interface. You can turn on “Separate mode” in Settings under Voice Mode.

A Push Toward Fully Multimodal AI

This update shows OpenAI leaning deeper into multimodal experiences. You could already talk to ChatGPT with voice, images, or video. Now, the reply format catches up. Although it is not reactive to live video like Google’s Gemini Live, the new ChatGPT Voice gives users richer and more expressive responses.