How to Use ChatGPT to Transcribe Audio Files

Step by step

1. 1 min

   Verify your ChatGPT Plus subscription

   Open ChatGPT and confirm you have an active Plus subscription. Audio file uploads require ChatGPT Plus or Team access. Free accounts cannot upload audio files for transcription.

   Note: The audio transcription feature uses the same Whisper model that powers OpenAI's dedicated Whisper API.
2. 1 min

   Prepare your audio file

   Ensure your audio file is under 25MB and in a supported format: MP3, MP4, MPEG, MPGA, M4A, WAV, or WEBM. For larger files, use audio editing software to compress or split the file into smaller segments.

   Note: Higher quality audio with minimal background noise produces more accurate transcriptions.
3. 1 min

   Upload the audio file

   Click the paperclip icon in the ChatGPT interface. Select your audio file from your device. Wait for the upload to complete — you'll see a preview of the file name in the chat input area.

4. 1 min

   Request transcription

   Type your transcription request in the message box. Use specific prompts like "Transcribe this audio file" or "Please provide a transcript with timestamps." Press Enter to send your request with the attached audio file.

   Note: Add specific formatting requests like "Include speaker labels" or "Format as a bulleted summary" for customized output.
5. 1 min

   Review and refine the transcript

   ChatGPT will process the audio and return a text transcript. Review the output for accuracy and formatting. Ask follow-up questions to correct errors, add formatting, or request specific edits to the transcription.

   Note: You can ask ChatGPT to summarize the transcript, extract key points, or format it as meeting notes without re-uploading the file.
6. 1 min

   Export or save the transcript

   Copy the transcript text from ChatGPT's response. Paste it into your preferred text editor, document processor, or note-taking app. Save the file in your desired format for future reference.

   Note: Use ChatGPT's code block formatting request to get plain text that's easier to copy without formatting issues.