ChatGPT adds real-time live dictation and removes recording limits
OpenAI's new update allows users to see text appear instantly as they speak, removing the 10-minute cap on voice inputs.
The short version
- Live Dictation displays recognized text on screen instantly as the user speaks.
- The previous 10-minute recording limit for voice inputs has been removed.
- Users can review and edit the converted text before sending it as a message.
- The update is available to all users, including those on the free plan.
- The feature is distinct from the interactive voice conversation mode.
Why it matters
This change makes it much easier to use voice for long-form content like meeting notes or drafting documents. By removing time constraints and showing text instantly, OpenAI is positioning voice as a primary way to input large amounts of data.
Real-time speech to text
OpenAI has officially introduced a new feature called Live Dictation to the ChatGPT platform. This tool functions as a real-time speech-to-text system that displays what a user says as they are speaking. It represents a shift from the previous method where users had to wait for a recording to finish before seeing the converted text.
The primary goal of this update is to improve the speed and stability of how the app handles speech recognition. By moving to a real-time streaming model, the system can now show recognized content on the screen immediately. This allows for a more fluid experience where the user can see the AI's interpretation of their voice in the moment.
As reported by 디지털투데이, the update focuses on making the transcription process feel more responsive. Users can interact with the input screen by tapping or clicking on the waveform. This action allows them to show or hide the text that is currently being converted from their voice into written words.
Removing time constraints
One of the most significant changes in this update is the removal of the 10-minute recording limit. Previously, users were restricted in how long they could record a single voice input. This new change allows for much longer voice entries without the need to stop and restart the recording process frequently.
While OpenAI has not stated a specific new maximum recording time or confirmed if the use is completely unlimited, the removal of the cap is a major shift. This change is specifically designed to help users who need to input long questions, detailed meeting notes, or complex ideas by voice. It removes the friction of having to manage short segments of speech for long-form content.
Atty Eleti, who manages OpenAI's voice features, noted on X that the move to real-time streaming improved message-sending speed and stability. By removing these time constraints, the company aims to make voice-based document writing much more convenient for the average user.
How the workflow functions
It is important to distinguish Live Dictation from the existing ChatGPT voice conversation mode. While the conversation mode is designed for back-and-forth dialogue with the AI, dictation is strictly for converting a user's speech into a text message. This means the user maintains more control over the final output before it is processed by the AI.
Once a user finishes speaking, the system provides a window to review the converted text. Users can then edit the transcription to correct any errors or refine the phrasing. After the text is finalized, the user can choose to send it to ChatGPT as a standard message for processing or further interaction.
This workflow is intended to be more accurate and reliable than previous versions. OpenAI previously launched a new speech-recognition model in June to improve accuracy across multiple languages, including Korean. That model reportedly achieved a word error rate that was at least 10 percent lower than the previous version.
Availability and rollout
OpenAI is rolling out the Live Dictation feature sequentially to its global user base. The update will be available across multiple platforms, including the web, desktop apps, and mobile applications for both iOS and Android. This ensures that users can access the new dictation tools regardless of their preferred device.
The update is not restricted to premium subscribers. OpenAI has confirmed that all ChatGPT plans, including those for free users, are eligible for this new feature. This broad availability suggests that OpenAI wants to make high-quality voice input a standard part of the user experience for everyone.
By combining improved speech-recognition accuracy with real-time processing, the company expects to see more users adopting voice for long message entry. The goal is to make the transition from spoken thought to written text as seamless as possible for people who prefer speaking over typing.
What to watch
- Whether OpenAI will eventually introduce a new hard limit on recording length.
- How the accuracy of the real-time streaming compares to the previous batch-processing method in noisy environments.
Try it
Open the ChatGPT app on your phone and try using the voice input to dictate a long paragraph of notes to see if the text appears instantly.
Sources and links
- Reporting디지털투데이: the article we sourced
First reported by 디지털투데이. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.