How do I read and copy a transcript?
Run a speech-to-text model with an audio or video file in its slot; the result opens in a transcript viewer with the text, its word count and a Copy button, and the .txt is saved in your project like any other file so you can download it.
Running speech-to-text
- Select a speech-to-text model, for example xAI STT v1 in the Grok Speech family.
- Drop the recording — an audio file or a video — into the reference slot.
- Press Generate. The estimate is per minute of the file's length.
The viewer
The result card opens the transcript viewer: the text as selectable, wrapped paragraphs, the word count in the header, and a Copy button that confirms with Copied. Very long transcripts are shown in full up to 2 MB of text.
Keeping it
The transcript is stored as a .txt file in your project, next to the source recording. Use the library's Download action to save it, rename it like any file, tag it, or send it to a teammate.
What you cannot do with it
A transcript is text, so it never appears in a reference picker: no model takes a text file as an input. To reuse the words, copy them.
Price
| Model | Starting price |
|---|---|
| xAI STT v1 | 4 credits per minute |
The charge is on the length of the audio you submit, rounded by the provider's own rule, and settles to the provider's real price when the result lands.
Frequently asked
What does transcription cost?
xAI STT v1 starts at 4 credits per minute of audio.
Can I transcribe a video?
Yes. Drop the video into the audio slot; its soundtrack is transcribed.
Can I use the transcript as a prompt?
Copy it and paste it into the prompt box, or into a text-to-speech model's script. It cannot be attached as a reference.