AI WORLD HQTalk to ChatGPT by voice on your phone while cooking or driving
Set up ChatGPT's voice mode in about a minute so you can brainstorm, settle an argument, or plan dinner hands-free.
In short: With the ChatGPT phone app you can have a spoken conversation with the assistant — ask out loud, hear the answer, follow up — hands-free. Setup takes about a minute; some of the most natural-sounding voices may need a paid plan.
By the end of this guide you'll have ChatGPT's voice mode working on your phone, ready to chat with you out loud while your hands are busy. This is for anyone who wants a hands-free way to brainstorm, settle a small argument, plan dinner, or just keep a conversation going without staring at a screen.
💡 Tip: tap a step’s number when you finish it — a green tick appears and your browser remembers how far you got.
- An iPhone or Android phone with the free ChatGPT app installed (App Store on iPhone, Google Play on Android). Tablet works too.
- A ChatGPT account — the free plan is enough to try voice mode, and sign-up takes about two minutes with any email address.
- About five minutes total: two for install and sign-in, three for your first conversation.
- A calm-ish first moment. Test voice mode somewhere quiet first (a kitchen counter, a parked car, a quiet room). Once you know it works, you can take it anywhere.
Install the ChatGPT app
Open your phone's app store — the App Store on iPhone, Google Play on Android — and search for "ChatGPT". The official app is published by OpenAI and shows a black square icon with a white, interlocking flower-like logo. Tap Install or Get, then wait for the download to finish.
If the store search is crowded with lookalikes, ignore anything that asks for payment upfront or has a different publisher name. The official one is free and clearly labeled "OpenAI" as the developer.
You'll know it worked when you see the black ChatGPT icon on your home screen and tapping it opens a screen that says "Welcome to ChatGPT."
Sign in or create a free account
When the app opens, you'll see a Sign up button and a Log in link at the bottom. If you already have a ChatGPT account, tap Log in and enter your email and password. If not, tap Sign up — you'll need an email address you can check (Gmail, Outlook, Yahoo, or any other works) plus your name and birthday.
Sign-up usually sends a verification code to your email. After that, you land directly on the free plan. The free plan covers basic voice mode; the most natural-sounding voices and longer voice sessions may need a paid plan like ChatGPT Plus — we'll get to that distinction in a moment.
You'll know it worked when the screen shows "How can I help today?" with a message box at the bottom.
Tap the microphone icon to start voice mode
At the bottom of the chat screen, next to the text box, look for a microphone icon 🎤 — or sometimes a headphones or sound-wave icon. Tap it once. The screen changes: the keyboard slides away and a circular animation appears, usually labeled "Listening…" with a pulsing dot in the middle.
If you can't spot a microphone on the main screen, look inside a small "⋯" or "+" button near the message box — some app versions tuck voice there. On Android it sometimes lives one tap deeper than on iPhone, so don't panic if it takes a second to find.
You'll know it worked when the screen shows a clear listening indicator and the text box is replaced by a voice prompt.
Ask your first spoken question
Once you see the listening indicator, talk in a normal voice, as if a friend were sitting across from you. Speak one clear sentence at a time, then pause for a moment. The app shows your words as text on the screen as it hears them, so you can check it understood you. When you stop talking, ChatGPT responds out loud through your phone's speaker.
If you're in a noisy kitchen, hold the phone a little closer to your mouth or step away from the blender. If you're in the car, use Bluetooth or a phone mount so you don't need to pick the device up. Keep your eyes where they need to be — on the road, on the cutting board — and let your ears do the work.
You'll know it worked when you hear ChatGPT's spoken answer come back through the speaker within a couple of seconds, in a voice that sounds like a real person rather than a robot.
Keep the conversation going naturally
Voice mode works like a real chat — you can interrupt, ask follow-ups, or change direction mid-answer. If ChatGPT starts rambling, just say "shorter please" or "give me the top three." If you want to switch topics entirely, just ask the new question; you don't need to say "new topic" first.
You can mute the speaker any time by tapping the speaker icon mid-conversation, or end the chat with the X or End button. When you end, voice mode stops listening and the regular text box returns. Your spoken conversation is automatically saved as text in your chat history, so you can scroll back later or copy a recipe into your notes.
You'll know it worked when you can fire off three or four follow-up questions in a row without tapping anything except maybe the mic once at the start.
Change the voice or stop the session
To swap the voice, look for a settings gear ⚙ or a small voice icon while you're in voice mode — a panel usually slides up with several voice options, each with a different tone or accent. Tap any of them to hear a short sample, then pick the one that sounds most natural to your ear. The voice you choose sticks as your default for future chats.
To finish, tap End or the X button. The screen goes back to normal chat mode, and your conversation appears in the sidebar with all the spoken words saved as text. Next time you want hands-free, open any chat and tap the mic again — it remembers your chosen voice.
You'll know it worked when the next voice session greets you in the voice you picked last time.
- Talking before the listening indicator appears. The app needs that half-second to start capturing audio cleanly. Wait for the pulsing circle or "Listening…" text, then speak — otherwise the first word or two often gets cut off.
- Expecting ChatGPT to browse live recipe sites or traffic reports on command. Voice mode can use general knowledge and recent training data, but it can't reliably pull from a live URL or a paywalled article while you're chatting. If the answer feels vague, finish what you're doing first, then ask for sources in text mode.
- Looking at the screen while driving. Voice mode is hands-free for talking and ears-free for listening, but your eyes still need to stay on the road. For anything visual — long answers, lists, links — pull over first, or hand the mic to a passenger and let them drive.
Open the ChatGPT app, tap the microphone once, and ask: "I'm about to cook with [whatever's in your fridge right now]. What can I make in twenty minutes?" Then just listen, follow up, and let your hands stay where they're needed.
❓ Quick questions
How long does this take?
About 6 minutes — the guide has 6 steps, and you can tick each one off as you go.
Which tool do I need?
This guide uses OpenAI ChatGPT — but the approach works very similarly in other AI assistants.
Do I need to prepare anything?
- An iPhone or Android phone with the free ChatGPT app installed (App Store on iPhone, Google Play on Android). Tablet works too.
- A ChatGPT account — the free plan is enough to try voice mode, and sign-up takes about two minutes with any email address.
- About five minutes total: two for install and sign-in, three for your first conversation.
- A calm-ish first moment. Test voice mode somewhere quiet first (a kitchen counter, a parked car, a quiet room). Once you know it works, you can take it anywhere.
What mistakes should I avoid?
- Talking before the listening indicator appears. The app needs that half-second to start capturing audio cleanly. Wait for the pulsing circle or "Listening…" text, then speak — otherwise the first word or two often gets cut off.
- Expecting ChatGPT to browse live recipe sites or traffic reports on command. Voice mode can use general knowledge and recent training data, but it can't reliably pull from a live URL or a paywalled article while you're chatting. If the answer feels vague, finish what you're doing first, then ask for sources in text mode.
- Looking at the screen while driving. Voice mode is hands-free for talking and ears-free for listening, but your eyes still need to stay on the road. For anything visual — long answers, lists, links — pull over first, or hand the mic to a passenger and let them drive.
Keep reading
✦ Original step-by-step guide by AI World HQ's AI editorial team. Written in plain language, reviewed for accuracy.
← Back to all stories