OpenAI is introducing an upgraded voice dictation feature for ChatGPT, allowing users to see their spoken words converted into text in real time while removing the previous 10-minute recording restriction.

The updated Live Dictation capability is being rolled out internationally across free and paid ChatGPT plans, including access through iOS, Android, desktop applications and web browsers.

Previously, users recording voice prompts had to wait until they finished speaking before reviewing the transcribed text. Longer recordings were also restricted by a 10-minute limit, making the feature less practical for extended conversations, detailed instructions and lengthy notes.

The latest update changes this process by displaying text as speech is recorded. Users can monitor the transcription, identify potential errors and adjust their wording without having to stop and restart the recording.

Removing the recording limit also provides greater flexibility for people who prefer speaking rather than typing, particularly when preparing lengthy documents, developing ideas or providing complex instructions to ChatGPT.

The feature could prove useful for professionals working remotely, students preparing study notes and anyone using voice input to capture information while multitasking.

Unlike conventional smartphone dictation tools, which generally convert speech into text within individual applications, ChatGPT’s Live Dictation operates directly within its conversational interface. This allows users to review spoken prompts before submitting them to the AI assistant.

OpenAI has also introduced an option to hide the live transcription while recording. By selecting the audio waveform displayed on screen, users can switch the text preview on or off, depending on whether they prefer to monitor their words or concentrate on speaking.

Although OpenAI describes the upgraded dictation system as faster and more reliable, the company has not released detailed figures comparing transcription accuracy or processing speeds with the previous version.

It has also not specified a new maximum recording duration, despite removing the earlier 10-minute restriction.

Privacy settings remain available for users concerned about sharing voice recordings. ChatGPT’s data controls allow users to manage whether audio recordings are shared and whether their conversations can be used to improve OpenAI’s models.

The update forms part of OpenAI’s continuing development of voice-based interaction, with the company seeking to make spoken communication a more practical alternative to traditional keyboard input.

Australian users will receive access progressively as the international rollout continues.

Once available, Live Dictation can be activated by selecting the microphone icon within ChatGPT and beginning to speak, with the resulting text appearing as the recording progresses.