OpenAI has recently removed the small microphone button that was previously positioned next to the send button on the ChatGPT iPad app. This change significantly affects how users can interact with the app using voice input.
Before, users could:
1. Dictate at their own pace – They could build a prompt gradually, stopping and resuming as needed.
2. Edit with voice – They could place the cursor anywhere in the text and dictate additional input, refining or expanding their prompt seamlessly.
3. Dictate in parts – Users could start a prompt with voice, then manually edit or continue dictating another part without being forced to submit immediately.
Now, after using the microphone once, it disappears and turns into a send button, removing the ability to dictate and edit incrementally. This forces users to complete their entire input in one go, which is frustrating, inefficient, and a step backward in usability—especially for those who rely on voice input due to accessibility needs.
This change feels like a regression rather than an improvement, and we urge OpenAI to bring back the separate microphone button that was independent of the send function.
9 Likes
THANK YOU.
I have been searching this issue for hours and made a whole account on this website for it!
ONLY SOLUTION I FOUND is that this is not happening when logged into other accounts … Still very frustrating because I have some very impactful conversations on my main account.
This must be a recent glitch. I’m on iPhone. But thank you again because you perfectly put into words what I’m going through.
1 Like
Same here! A majority of the time, I use their microphone to dictate lots of information into text, which I can either send, or can revise, add onto using the microphone again, then sending.
I just noticed this morning that the microphone now disappears after its first use in a message and is instead replaced by either the Send or Voice Mode buttons (Voice Mode being obviously different than the Microphone button for voice-to-text Whisper transcription). Therefore, I cannot add onto existing text transcribed for the prompt I am still compiling.
I hope they re-introduce the permanence of the microphone button. It is highly accurate and would you let you easily add more text via speech.
4 Likes
I noticed this too since yesterday it’s extremely annoying us. This was one of the best features when reading at night.
3 Likes
Subject: Concern Regarding Removal of Voice-to-Text Functionality
Dear OpenAI Team,
I am writing to express my concern regarding the recent removal of the voice-to-text functionality. Many users, including paying customers, have subscribed to OpenAI for specific features, and the sudden removal of this capability without prior notification is disappointing and unfair to customers who rely on it.
This decision negatively impacts those who have been using the feature for their workflow, accessibility, and efficiency. At the very least, users should have been informed in advance and given an explanation for the change.
I request OpenAI to reconsider reinstating this functionality or, at the very least, provide clarity on the decision. If such significant changes continue without transparency, it raises concerns about customer rights.
I appreciate your attention to this matter and look forward to your response.
Best regards,
6 Likes
It is a big step backward, and will push more customers to depart aand move to competition - claude already has it and it works across all devices
3 Likes
Yeah this is a big deal for me as well it’s the way better than any IOS or Android TTS by far and they’re just going to give this up?
1 Like
Hi all — just adding my voice here to echo the concerns already raised. I’m another frustrated user who relied a lot on the original voice input features in the ChatGPT app.
The removal of the standalone microphone button on iPad, along with the changes to voice dictation — especially the inability to review or edit the transcription before sending — has seriously impacted both usability and accessibility. What used to be a smooth, flexible experience now feels limiting and unnecessarily rigid.
I used to build prompts gradually, edit by voice, and take my time — that was a huge part of my workflow. Now, that control is gone. It’s not just about convenience either — for users like me who have an accent and depend on voice input, this change has made things significantly harder.
Please, OpenAI: bring back the separate mic button and restore the option to preview and edit transcribed text before it gets sent. These features made a big difference and their removal has taken away a lot of what made the app great.
Thanks to everyone who’s spoken up already — really hope the team is paying attention.