Audio to text

Transcribe up to 1 hour of speech with local AI, without external services. Auto-detect or choose a language; check the text before using it. TXT output.

Private processing

Choose files

No file selected

One audio file, up to 500 MB and 1 hour. Processing limit 1 hour.

Spoken language

Files are used only for the requested processing. Queued and running jobs expire within 2 hours; results within 30 minutes or when you reset the tool. Check the result before use. Privacy notice.

I confirm that I have the necessary rights and permissions to process the files and accept the Terms of use.

Result

Choose files and options to prepare the download.

Download result

We use your files only for the result you request. No registration.

We count which tools are used most, without including entered values or file contents in the statistics. Privacy

How to use the tool

Initial form values
Parameter	Value
Spoken language	Auto detect

Practical example

Usage scenario

Input: WAV / MP3; en
Result: TXT (UTF-8)

Limits and checks

Files are used only for the requested processing. Queued and running jobs expire within 2 hours; results within 30 minutes or when you reset the tool. Check the result before use.

Sources and further reading

Faster Whisper · MIT

Frequently asked questions

How long do files remain on the server?

Queued or running jobs expire within 2 hours; results expire after browser receipt, reset or 30 minutes. Periodic cleanup can take a further 5 minutes. Any backup copies follow the policy described in the privacy notice.

PRACTICAL GUIDES

Turn audio into a transcript you have actually checked

A fluent transcript can still contain a mistake that reverses the meaning. Finish the job by comparing the text with the recording, rather than accepting the first automatic result.

Follow the steps