Bulk audio transcription
Transcribe a whole folder of recordings in one job
Select up to 100 audio or video files, or upload one ZIP of them, and every file is transcribed in full with timestamps counted from the start of that file. You download one merged document with a heading per file, plus a ZIP with a separate TXT and SRT for each recording. A batch of up to 15 minutes in total is free in full. A bigger batch is $5, paid once, for up to 6 hours of audio in total.
- Up to 100 files in one job, or a ZIP with folders inside, 4 GB of uploads in total
- One price for the batch: $5 covers up to 6 hours of audio in total (12 hours on Unlimited)
- Files handled in name order, folder path included, each with timestamps from 0:00
- An identical file uploaded twice is counted once
- Optional translation under every paragraph of every file
- One merged Word, PDF or TXT document, and a ZIP with a TXT and an SRT per file
- Up to 15 minutes in total free in full, first part of the first file free · whole batch $5, one payment
What you get back
Every recording written out word for word, kept apart by file, so you always know which text came from which recording.
- A merged Word, PDF or TXT document that opens with a list of the files and their durations.
- Each file under its own heading in that document, in name order.
- A ZIP with a numbered TXT and SRT for every file, such as "01 interview.txt" and "01 interview.srt", plus the merged text.
- A timestamp such as [12:40] at the start of every paragraph, counted from the start of that file.
- Speaker changes marked as Speaker 1, Speaker 2 when they are clear; unclear words marked [inaudible].
- If you choose a translation, it sits under each paragraph, with the original first.
Doing it in a chatbot, and with this tool
A chat assistant can handle one short, clear clip. Forty recordings are a different job: each one has to be written out in full and kept apart from the others.
| What you need | A chatbot | This tool |
|---|---|---|
| One short voice memo | Often fine, if your plan accepts audio | Also fine, and free in full up to 15 minutes |
| Thirty files from one project | One upload at a time, results pasted together by hand | Up to 100 files or a ZIP in a single job |
| Long recordings in the batch | Upload limits, shortened or unfinished text | Every file cut into parts and written out in full |
| Knowing which text belongs to which file | Easy to mix up across a long chat | A heading per file in the merged document, a separate file per recording in the ZIP |
| Times you can jump back to | No reliable timestamps | A timestamp on every paragraph, from the start of each file |
| Files to keep or import | Text in a chat window | Word, PDF, TXT and a ZIP of TXT and SRT files |
How to transcribe multiple audio files at once
-
1
Upload the files or a ZIP
Select up to 100 audio or video files (MP3, M4A, WAV, OPUS, AMR, MP4, MOV and more), or drop in one ZIP. Folders inside the ZIP are fine.
-
2
Set language and translation
Pick the spoken language if every file shares it, or keep Auto-detect for a mixed batch. Add a translation language only if you want one under each paragraph.
-
3
Check, pay once, download
Read the first part of the first file free on screen. Unlock the batch for $5, then download the merged document or the ZIP with a file per recording.
Why a batch of recordings is slow in a chat window
Chat assistants work one conversation at a time. Even when your plan accepts audio, you upload a file, wait, copy the result, and start again for the next one, and long files run into the upload limit or come back shortened. With twenty or fifty recordings, most of the time goes into handling files rather than reading them, and it is easy to paste a transcript under the wrong name.
Here the whole batch is one job. Each file is cut into parts of up to ten minutes, every part is written out word for word by an AI speech model, and the parts are joined again per file, with the times counted from the start of that file.
What people send in bulk
- Research interviews and focus groups for a thesis or a study
- Recorded customer or sales calls to review
- A semester of short lectures or a training course split into many files
- Voice memos and field recordings from a phone
- Podcast or radio archives that were never transcribed
- A family archive of cassette or voicemail recordings converted to MP3
How the files and the total are counted
Files are processed in name order, and inside a ZIP the folder path counts too, so "Week 1/Lecture 2.mp3" comes before "Week 2/Lecture 1.mp3". Numbers in names are sorted as numbers, so "Part 2" comes before "Part 10". Put a number or a date at the front of each name if the order matters. Non-audio files in a ZIP, hidden files and the "__MACOSX" copies that a Mac adds are skipped. An identical file that appears twice is transcribed only once.
The price follows the total length, not the number of files. One payment of $5 covers up to 6 hours of audio across the whole batch, and Unlimited covers up to 12 hours in total per job. Audio past that total is not included, so split a larger archive into more than one job.
Accuracy and limits
The transcripts are made by an AI speech model. Clear speech comes out well; background noise, people talking over each other, phone-quality lines and far microphones cause mistakes, so check names and numbers. Speaker labels are added when the change of speaker is clear, and on a long file the same person can get a different number in a later part: use Find and replace on screen before you download.
Bulk jobs take uploaded files only, not links, so download recordings from cloud storage first. A single file can be up to 2 GB. The text is a draft to check, not a certified transcript.
Bulk transcription for specific jobs
Pages for common batches, such as a ZIP of recordings or a podcast back catalog.
Transcribe multiple audio files at once
Select up to 100 audio or video files at once and get every file transcribed in full, with timestamps counted...
Read more →Transcribe a ZIP file of recordings
Upload one .zip with up to 100 audio or video files inside, folders and all, and each recording is transcribed...
Read more →Transcribe all your podcast episodes
Upload your episode files, up to 100 per job, and get every episode transcribed word for word with timestamps...
Read more →Transcribe a study’s interviews in bulk
Upload the 10 or 20 interviews of a study together and get each one transcribed word for word as its own TXT f...
Read more →What it costs
Small batches, first part
Free
A batch of up to 15 minutes in total is free in full, downloads included, no account. For a bigger batch, the first part of the first file is shown on screen so you can judge the quality.
Whole batch
$5
One payment for up to 100 files and 6 hours of audio in total. Merged Word, PDF, TXT and the ZIP of TXT and SRT files. No subscription.
Unlimited
$19 / week
Every tool on the site, batches up to 12 hours in total per job. Renews weekly, cancel any time.
Questions and answers
How many files can I upload at once?
Up to 100 audio or video files in one job, or one ZIP that contains them, with 4 GB of uploads in total. A single file can be up to 2 GB.
Can ChatGPT transcribe multiple audio files at once?
It can work through a few short clips, one after another. With many or long files, upload limits and the length of one reply tend to cut transcripts short, timestamps are not reliable, and you still paste the results together yourself.
Is it priced per file?
No. One payment of $5 covers the whole batch, up to 6 hours of audio in total, whether that is three long recordings or many short ones. A batch of up to 15 minutes in total is free in full.
What do the downloads look like?
A merged Word, PDF or TXT document that starts with a list of the files and their durations, then each file under its own heading. The ZIP holds a numbered TXT and SRT for every file and the merged text.
Can the files be in different languages?
Yes. Leave the spoken language on Auto-detect and each file is written in the language spoken in it, including mixed speech. A translation, if you choose one, goes under every paragraph.
What happens if I upload the same file twice?
Identical files are recognized by their content and transcribed once, so a duplicate does not use up your total.
Can I paste links instead of uploading?
Not for bulk jobs. Download the recordings first, or zip a folder and upload the ZIP.
Are my recordings private?
Only you can open your job. The files are sent to an AI model provider only to be transcribed and are not used to train models. Jobs started without an account are deleted after 7 days, and you can delete a job yourself at any time.
Updated October 25, 2026 · APPFACT STUDIO LTD