Audio to text, ready for the next Agent step

Choose a recording, transcribe it into text and subtitle files, then pass the transcript to a Vecbase Agent as a file attachment for summaries, minutes, translation, or interview analysis.

Drop audio or video here

MP3, M4A, WAV, FLAC, OGG, WEBM, AAC, MP4, MOV · up to 25 MB

Three steps to a finished result

  1. 1

    Choose the recording

    Add one audio or video file up to 25 MB. If you are not signed in, the file stays in the browser until login completes.

  2. 2

    Generate transcript files

    Vecbase creates plain text, SRT subtitles, and a segment JSON file that can be downloaded or used by an Agent.

  3. 3

    Continue with an Agent

    Start a follow-up Agent with the transcript attached as a file, so long recordings do not become oversized prompts.

Why the result is easier to use

Outputs the files teams need

TXT works for reading and search, SRT works for subtitles, and segments.json keeps structured timing for downstream tools.

Keeps follow-up work file-based

The Agent receives the transcript as an attached file and can summarize, translate, or analyze it without stuffing the entire transcript into a prompt.

Clear limits before upload

The first version uses a 25 MB file cap and returns a visible error when a format or transcription request is not accepted.

Questions before you start

What files can I upload?

Use MP3, M4A, WAV, FLAC, OGG, WEBM, AAC, MP4, or MOV up to 25 MB. Compressed audio is better for long recordings.

What do I get after transcription?

You can download transcript.txt, transcript.srt, and segments.json. If the upstream model does not return exact segment timings, Vecbase creates reviewable approximate cues from the transcript and duration.

Can an Agent summarize this transcript?

Yes. After transcription, choose a follow-up such as summary, meeting minutes, translation, subtitles, or interview analysis. The transcript is attached as a file reference.

Does this replace human review?

No. Names, jargon, noisy sections, and overlapping speakers still need review before publishing captions or using the transcript as evidence.

Vecbase

Ready to start audio transcription?

For audio transcription, add an audio or video file, explain what matters, and let the Vecbase Agent prepare TXT, SRT, and segment files you can review and refine.