Condenser microphone used to record audio that can be uploaded to ChatGPT for a transcript

ChatGPT Won’t Accept Your Audio File? Who Can Upload Recordings, the 512 MB Limit, and Why MP4s Fail (2026)

Short answer: ChatGPT can now transcribe and summarize audio files you upload, but only on paid plans. Free accounts don’t get it. Files can be up to 512 MB. MP3, M4A, WAV, FLAC and a few other formats work. The most common failure is a video file. An MP4 or WebM with a video track gets rejected even if you only care about the sound, so strip the video first. Very long recordings can still time out.

You drag a meeting recording into ChatGPT, and nothing useful happens. Maybe the attach button won’t take it. Maybe it uploads and then errors out.

OpenAI added audio uploads to ChatGPT on October 6, 2026. It’s handy: drop in a recording and ask for a transcript, notes, or a follow-up email. But the rules on who gets it and which files work are stricter than the announcement made them sound.

This guide walks through each reason an audio upload fails. It also covers how long a recording actually fits under the 512 MB cap.

In This Fix

Who can upload audio to ChatGPT?

Audio uploads in ChatGPT are available on paid ChatGPT subscriptions and workspaces, including Enterprise, and are not available on the Free plan, according to OpenAI’s Uploading files and audio to ChatGPT help page as updated in October 2026.

That’s the first thing to rule out. A Free account can still upload documents and images, so the paperclip button is there. It just won’t take an audio file.

Paid users can still miss out. OpenAI says availability “may vary by workspace settings, region, client version, and selected model.” In practice that gives you three more checks:

  • Update the app. An older ChatGPT app may not offer audio at all.
  • Ask your admin. On Business or Enterprise, a workspace admin can switch features off.
  • Try another model. If one model in the picker won’t take the file, switch and try again.

Don’t mix this up with Voice mode or the Record button. Those are separate features with their own limits. This guide is only about attaching an existing audio file.

Why won’t ChatGPT accept my audio file?

ChatGPT rejects audio files that are in an unsupported format, that are video files, or that lack a valid, decodable audio stream; OpenAI’s October 2026 help page lists WAV, MP3/MPEG, OGG/OGA, audio-only WebM, PCM, FLAC, AAC, M4A, and audio-only MP4 as the supported audio formats.

When I checked that list against the files people actually have lying around, I found that the gap is almost always video. Here’s how common recordings line up.

Where the recording came from Typical file Will ChatGPT take it? What to do
iPhone Voice Memos M4A Yes Share the memo to Files, then attach it
Podcast or music export MP3, WAV, FLAC Yes Watch the 512 MB cap on WAV
Zoom or Teams meeting video MP4 with video No Use the audio-only file, or strip the video
Phone camera clip MP4 or MOV with video No Convert to M4A or MP3 first
Browser screen recorder WebM with video No Export audio only
Half-finished or interrupted recording Any format Often no Open it in a player first; if it won’t play, it won’t upload

Supported formats per OpenAI’s help page, checked October 13, 2026. Typical file types are the usual defaults for each app and can differ by settings.

That last row is easy to miss. OpenAI says the file “must contain a valid, decodable audio stream.” A recording that cut off when your phone died can have a broken header. ChatGPT can’t read it.

How do you upload a Zoom or phone video as audio?

ChatGPT does not accept MP4 or WebM files identified as video for audio uploads, so a Zoom recording or phone video has to be turned into an audio-only file such as M4A or MP3 before you attach it.

Method 1: Grab the audio file your app already made

  1. Check your recording folder. Zoom’s local recordings usually save an audio-only M4A next to the MP4. Use that one.
  2. Download the audio version from cloud recordings when the service offers one. Many meeting tools list “Audio only” as a separate download.
  3. Attach the M4A to a new ChatGPT chat and ask for a transcript.

Method 2: Strip the video yourself

  1. Install ffmpeg, the free command-line tool documented at ffmpeg.org.
  2. Run ffmpeg -i meeting.mp4 -vn -c:a copy meeting.m4a. The -vn drops the video. -c:a copy keeps the original audio without re-encoding, so it’s fast.
  3. If that errors, re-encode instead: ffmpeg -i meeting.mp4 -vn -b:a 64k meeting.mp3.
  4. Upload the new file. It’ll also be far smaller than the video.

Not a command-line person? VLC’s Convert/Save option can do the same thing with an “Audio” profile. It takes a few more clicks.

Person with headphones checking a ChatGPT meeting transcript against the original recording
ChatGPT audio uploads cap out at 512 MB per file, and OpenAI warns that speaker identification in transcripts may be unreliable (OpenAI Help Center, October 2026).

How long a recording fits under the 512 MB limit?

A 512 MB ChatGPT audio upload holds roughly 9 hours of 128 kbps MP3 audio but only about 50 minutes of uncompressed CD-quality stereo WAV, so the file format matters far more than the recording length.

OpenAI states the 512 MB cap but doesn’t translate it into minutes. Here’s my own math, using standard bitrates:

Format and quality Data per second Approx. length in 512 MB
WAV, 16-bit 44.1 kHz stereo ~176 KB ~50 minutes
WAV, 16-bit 16 kHz mono (voice) ~32 KB ~4.6 hours
MP3 at 128 kbps 16 KB ~9 hours
M4A/AAC or MP3 at 64 kbps 8 KB ~18 hours

Calculated from bitrate alone. Real files carry a little overhead, so leave some margin.

Fitting the size limit doesn’t mean it’ll finish. OpenAI says longer recordings “may be processed in smaller sections when Data Analysis is available,” that processing is best-effort, and that “very long recordings may time out.”

So if a three-hour board meeting keeps failing, split it. Three one-hour files usually beat one big one. With ffmpeg: ffmpeg -i long.m4a -f segment -segment_time 3600 -c copy part%02d.m4a.

Getting “upload limit reached” when you’ve barely uploaded anything?

ChatGPT’s upload limit is 80 files every 3 hours on paid plans, plus a 25 GB storage cap per user, and OpenAI’s help page notes that failed upload attempts can sometimes count toward the upload-rate limit.

That last part explains a lot of confused support threads. Retry a bad video file ten times and you’ve burned ten slots without a single success.

  1. Confirm the account. Make sure you’re signed in to the paid account, not a personal Free one in another browser profile.
  2. Check storage in Settings, then Storage. Old uploads saved in Library count toward the 25 GB cap.
  3. Check OpenAI Status for upload incidents before you assume it’s your file.
  4. Wait out the window. ChatGPT doesn’t show how much of the rolling 3-hour quota you’ve used.

Can you trust ChatGPT’s transcript?

ChatGPT transcripts of uploaded audio may contain errors and its speaker identification may be unreliable, OpenAI warns in its October 2026 help documentation, and transcription quality can vary by language.

Use it for drafts and summaries. Be careful with quotes.

Reading OpenAI’s notes, I noticed that the warning covers exactly the parts people lean on most. Who said what, names, and numbers. For a meeting recap, that’s usually fine. For a legal record, an interview you’ll publish, or medical notes, check each key line against the recording.

One habit that helps: ask ChatGPT to mark timestamps next to decisions and numbers. Then you only re-listen to those spots instead of the whole file.

If you’d rather get the plain text without charts and widgets around it, our guide to getting plain-text answers after the GPT-6 Intelligent UI update covers the prompt that works.

Tips

  • Record voice at 64 kbps mono if the app lets you. It’s plenty for speech and keeps files small.
  • Name speakers at the start of the chat (“Speaker 1 is Dana, the host”). It gives ChatGPT something to anchor on.
  • Ask for the transcript first, then the summary, in two messages. You can check the transcript before trusting the summary built on it.
  • Delete the file from Library after you’re done if the recording is sensitive. Deleting the chat alone doesn’t remove it.

Warnings

  • Get consent before recording calls. Some U.S. states require every party to agree, and that applies no matter what tool transcribes it.
  • Don’t upload recordings with patient, client, or legal details on a personal plan without checking your data settings. Consumer chats can be used for training unless you turn that off.
  • Don’t retry a failing file over and over. Failed attempts can eat your upload quota.
  • The 25 MB limit you’ll see quoted online is for OpenAI’s API transcription endpoint, not for ChatGPT itself.

FAQ

Can ChatGPT Free transcribe audio files?

No. OpenAI’s help page says audio uploads are available on paid ChatGPT subscriptions and workspaces and are not available on the Free plan at this time. Free users can still upload documents and images.

Is the ChatGPT audio upload limit 25 MB or 512 MB?

For ChatGPT, it is 512 MB per audio file. The 25 MB figure applies to OpenAI’s legacy whisper-1 transcription API, which is a separate developer product with its own limits.

Does deleting the chat delete my uploaded recording?

Not always. OpenAI says deleting a chat does not delete a file that remains saved in Library. Open Library, select the file, and delete it there to remove the saved copy.

Sources

📌 Hub guide: For every fix, buying decision, and work-from-anywhere setup in one place — see the Tech & Digital Hub.

Fact-checked against OpenAI’s Help Center documentation as of October 13, 2026. Plan availability and limits can change; OpenAI may lower upload limits during peak hours.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *