OpenAI has added a Live Dictation feature to ChatGPT that transcribes spoken input in real time and eliminates the previous 10‑minute recording cap, available now on web, desktop, iOS and Android.
ChatGPT now supports Live Dictation, a mode that converts spoken words to text as the user talks instead of waiting for the recording to finish. Users can tap the audio waveform to view the live transcript and edit the generated text before submitting the prompt.
The update removes the earlier 10‑minute limit on voice input, allowing longer sessions without interruption. OpenAI says the new engine is also more reliable, building on a prior overhaul of the two‑way conversational voice mode introduced earlier this year.

Live Dictation is being rolled out globally across all ChatGPT subscription tiers, including free, Plus, Enterprise, Edu, Business and Pro plans, and is accessible through the web interface, desktop application, and mobile apps for iOS and Android.
The feature complements OpenAI's existing "record mode," which remains in beta and is limited to paid workspaces. Record mode focuses on transcribing longer voice notes and meetings, while Live Dictation offers instant, on‑the‑fly transcription for everyday queries.
TechTimes Commentary
TechTimes notes that real‑time transcription removes a friction point for users who rely on voice input, potentially expanding ChatGPT's appeal in mobile and accessibility contexts. By eliminating the 10‑minute cap, OpenAI positions the service for longer, more natural interactions, which could drive higher engagement across its paid tiers. Competitors offering similar dictation capabilities may need to accelerate their own updates to keep pace. The broader rollout suggests OpenAI is standardising voice features as a core part of its conversational AI suite.







