Skip to content
Theme
OnDevice Tools

AIChatbots & Assistants

ChatGPT Live Dictation transcribes speech as users talk

OpenAI has started rolling out ChatGPT Live Dictation, a feature that turns speech into text while a person is still talking. The update is available globally on all ChatGPT plans and removes the previous 10-minute recording limit.

Illustration of a ChatGPT dictation screen turning live speech into editable text

Key takeaways

  • Live Dictation turns speech into text while the user is still speaking.
  • It is rolling out globally on all ChatGPT plans across web, desktop, iOS and Android.
  • OpenAI has removed the 10-minute recording limit for voice dictation.
  • Users can reveal the live transcript from the audio waveform and edit it in the prompt box.
In this article
  1. Live transcription arrives in ChatGPT
  2. How the dictation flow works
  3. Longer recordings and wider access
  4. Place within ChatGPT voice updates
  5. Why the update matters

Live transcription arrives in ChatGPT

OpenAI has started rolling out ChatGPT Live Dictation, according to The Verge. The feature is designed to transcribe audio prompts faster. Instead of waiting for a person to finish recording, it produces text while the person is speaking. The Verge reported the rollout on 19 November 2026.

This changes the basic dictation flow in ChatGPT. Previously, a spoken prompt would be recorded first and then converted into text. With Live Dictation, the transcript appears during the recording. That makes the process more immediate.

The feature is part of ChatGPT's voice input tools. It does not change the typed prompt box, but it gives users a new way to fill it. The result can be reviewed before the prompt is sent.

The feature is not described as a replacement for typed input. It is an additional way to create prompts using the microphone. The update is therefore focused on making spoken prompts easier to create and review.

How the dictation flow works

Illustration: microphone, AI, settings, ChatGPT logo and the labels ChatGPT, Live Dictation, OpenAI

The Verge says users can tap the waveform or click it while dictating. This shows the live transcript produced by the AI. The transcript reflects what ChatGPT is hearing as the user speaks.

Once the recording is finished, the user can edit the text in the prompt box. This allows corrections before the prompt is submitted. The prompt box therefore remains the final place where the dictated text is prepared.

The same dictation feature can also be used to request a transcription of voice messages. It can also create bullet points from them. This gives the feature uses beyond simple command input.

The live transcript is generated by the AI, according to The Verge. This means the text is produced as part of the dictation process, not added later by the user.

Longer recordings and wider access

OpenAI has lifted the 10-minute cap on voice dictation in ChatGPT. The Verge reports that this change is part of the same update. The company also says the feature is becoming more reliable.

ChatGPT Live Dictation is rolling out globally. It is available on all ChatGPT plans. It works on the web, in the desktop app, and on iOS and Android.

The broad rollout means the feature is not limited to one platform. It also means users on different ChatGPT plans can access it, according to the report. The report describes record mode as beta, but does not give Live Dictation that label.

The update applies to voice dictation in the prompt box. It is distinct from the record mode that remains limited to certain workspaces. The record mode remains a separate feature with its own access limits.

Place within ChatGPT voice updates

The update comes after an earlier revamp of ChatGPT's conversational voice mode this year. That earlier work focused on spoken conversation with the assistant. Live Dictation is aimed at turning speech into text for a prompt.

OpenAI introduced a record mode last year for voice notes and meetings. That feature is still in beta. It is limited to Plus, Enterprise, Edu, Business, and Pro workspaces.

Live Dictation is separate from that record mode. It is broader in availability because it is reaching all ChatGPT plans. The two features show different approaches to voice input: one for conversation, one for transcription.

The Verge presents Live Dictation as the latest step in OpenAI's voice work. It builds on earlier changes while focusing on a simpler task: turning speech into editable text.

Why the update matters

The main benefit described by The Verge is speed. Live Dictation reduces the wait between speaking and seeing text. This may make voice prompts feel more responsive.

The removal of the 10-minute limit may make the feature more useful for longer spoken inputs. The report does not provide benchmark results or independent accuracy tests. Any claim about reliability comes from OpenAI.

For users, the practical point is that ChatGPT now offers a live dictation option across major platforms. Users can speak, watch the transcript appear, and edit the result before sending it. The feature is available globally on all plans, according to The Verge.

The rollout does not include new pricing details in the report. It also does not list any hardware requirements beyond the supported apps and platforms. The source does not mention any changes to ChatGPT's conversation mode in this rollout.

Sources

  1. ChatGPT can transcribe as you speak now with Live DictationThe Verge

Tags

Spotted a mistake? See our corrections policy. Read how News by OnDevice works in our editorial policy.