Can a quote be checked after the transcript is made?
Yes. Each line holds the time it starts and ends, so the sentence in the transcript can be compared with the recording in a second, and the journalist fixes a wrong name in the editor.
Transcribe Pashto audio to text: a Pashto interview needs its own engine, ours writes down 99+ languages. AssemblyAI lists Pashto among 99 languages.
Pashto audio needs an engine built for Pashto, and ours is not one: it transcribes 99+ languages and no more. AssemblyAI lists Pashto among the 99 languages it publishes (assemblyai.com/languages, read 22 October 2026). The interview goes there; the 99+ languages part comes to us.
An interview for a newsroom is one file with two jobs. The words have to be right, because a quote is going out under a byline, and the timecode has to point at the place in the recording where the words were said, so that any sentence can be checked in a second. Both jobs are easier when the language of the file is one the engine knows.
| What | Value |
|---|---|
| Pashto speech | on our list — our engine has 99+ languages |
| A service that lists Pashto | AssemblyAI lists 99 languages, Pashto among them |
| Length of one file | up to 30 minutes on the free plan, no minute limit on Pro |
| Free plan | $0: 3 files a day, up to 30 minutes per file, no card |
| Paid plan | $10 a month paid yearly ($120 a year), or $20 a month paid monthly |
A source who answers in Pashto and a journalist who asks in English give two recordings, or one file with a switch in the middle. Find the switch.
AssemblyAI names Pashto in the list of 99 languages it publishes, read on 22 October 2026. The address of that page is in the sources below.
The widget takes up to 30 minutes, and the speakers can be named on request, so the questions and the answers do not merge.
Every line carries the time it starts and ends. That is the moment to press play: a quote is one search and one listen away from being verified.
Keep the original recording next to the transcript.
Check the names in the first minute of the interview.
Export DOCX for the editor and TXT for the archive.
Ask for a verbatim transcript, not a cleaned one.
Yes. Each line holds the time it starts and ends, so the sentence in the transcript can be compared with the recording in a second, and the journalist fixes a wrong name in the editor.
No. Our engine covers 99+ languages. For a Pashto interview, AssemblyAI publishes Pashto among the 99 languages on its language page, read on 22 October 2026.
up to 30 minutes for one file, on every plan. A longer interview has to be cut into parts before the upload.
The free plan is $0 with three files a day and up to 30 minutes per file, no card. Pro is $10 a month paid yearly, $20 a month paid monthly.
No. The audio and the transcript are never used to train models for other customers and never sold. The privacy page says what leaves the server.
Which languages our engine takes, with the code of each and whether the language is detected automatically.
Transcribe Mongolian audio to textThe same question for another language.
Transcribe for journalistsHow a newsroom uses a transcript.
Transcribe English audio to textThe page about the language we do support.
Speakers by name, a time on every line, export to TXT, SRT, VTT, CSV and Markdown.