20:01Machine-translated from Azerbaijani by AIRead in Azerbaijani

OpenAI Introduces Live Speech-to-Text Feature in ChatGPT

OpenAI is rolling out “Live Dictation,” a feature for ChatGPT that converts speech into text in real time. Previously, dictation was converted to text only after the audio recording ended. Now, the text appears on the screen as the user speaks. The new feature is being made available worldwide on all ChatGPT plans. It works on the web version, the desktop app, and the iOS and Android apps.

During dictation, users can tap or click the waveform icon to see the generated text live. When they finish speaking, the text appears in the prompt box. It can be edited before being sent. This makes it easier to correct misrecognized words and prepare longer voice prompts.

With the same update, OpenAI has also removed the 10-minute recording limit for voice dictation. Users can now record long stretches of speech without dividing them into segments. According to the company, the dictation feature now works more reliably. However, its accuracy level and the languages it supports have not been disclosed.

ChatGPT’s interactive voice conversation mode was updated earlier this year. Last year, OpenAI also announced “record mode,” a feature that converts voice notes and meetings into text. Still in testing, the feature is available only in Plus, Enterprise, Edu, Business, and Pro workspaces. Standard dictation can be used to convert a voice message into text or divide that text into short sections.