Loading...
ChatGPT now transcribes your words live as you speak
Live Dictation is available on all platforms

ChatGPT now transcribes your words live as you speak

Nov 20, 2026
10:30 am

What's the story

OpenAI has introduced a new feature called Live Dictation for its ChatGPT service. The update allows the AI to transcribe audio prompts in real time, instead of waiting for the user to finish speaking. The feature is now available across all platforms including web, desktop app, and iOS and Android apps. Along with faster transcription, OpenAI has removed the 10-minute recording limit on voice dictation in ChatGPT.

Feature details

Making AI interactions more natural

The Live Dictation feature is part of OpenAI's larger effort to make interactions with AI more natural.

This comes after the launch of a voice model called GPT-Live-1 in July, which mimicked human speech patterns like spontaneous laughter and mid-sentence pauses.

An OpenAI researcher had described GPT-Live-1 as "one step closer to a truly accessible AGI, a world where talking to AI actually starts to feel like a real conversation."

User behavior

Users may overshare with AI

While OpenAI isn't explicitly marketing the new ChatGPT update as a "companion," it could lead to users oversharing.

A 2025 study by Danish audio company Jabra and the London School of Economics found that trust in AI increased by an average of 33% when using voice commands over text prompts.

The study involved 171 knowledge workers and suggests people may be more open with technology when communicating verbally.

ADVERTISEMENT

Data training

Audio data could be used to improve models

OpenAI has also hinted that it could use the data from Live Dictation to train its models.

"If you are a consumer user and have chosen to share audio to improve our models, we may train on audio from dictation," the company says on its website.

However, users can disable this sharing by going into Settings in ChatGPT and turning off "Improve the model for everyone."

ADVERTISEMENT