AI news · October 8, 2026
ChatGPT adds audio upload and meeting-note workflows for paid users
Paid ChatGPT users can upload audio up to 512MB, ask questions about a recording, and turn a meeting or lecture into notes. Free-plan audio support is not included.
- maximum audio upload size
- 512MB
- listed audio formats
- 8
- release-notes date
- Oct 6
OpenAI added audio uploads to ChatGPT on October 6 for paid users and workspaces. A user can upload a recording, get a transcript, ask questions, summarize the discussion, and produce structured notes or a follow-up email. OpenAI's file guidance lists a 512MB audio limit and support for common formats including WAV, MP3, OGG, WebM, PCM, FLAC, AAC, and M4A. Audio upload is not available on the Free plan, and the guidance warns that transcription can contain errors.
This makes ChatGPT a direct competitor to meeting-note tools for small teams, interviews, lectures, and research calls. The useful test is not whether a transcript appears; it is whether names, decisions, owners, dates, and unresolved points survive the conversion. Teams should compare a generated note with the recording, set a review owner, and avoid treating an unverified transcript as the system of record for a customer or legal decision.
What you can do with it
Try one short meeting recording in a paid workspace. Check speaker names, numbers, decisions, and action owners against the source audio. Save the transcript and final note separately, then measure editing time against your current meeting tool before moving the workflow.
Our take
Audio upload is a sensible distribution move because it meets users where recordings already exist. It can replace a simple note-taking step for low-risk meetings. The transcription warning matters: teams still need review before a generated note becomes an operational record.
Links Try today
Source: OpenAI ↗ — Made With Models writes the brief; the reporting is theirs.