About Audio & Video to Text
Typing out a long recording by hand takes hours: a lecture, a meeting, an interview, a podcast or a long voice note. This tool turns audio and video into written text with a timing for every line, gives you an SRT subtitle file ready for YouTube, TikTok and video editors, and then translates the text into another language with the same timings, or summarizes it into bullet points.
There are two modes. «Fast» runs entirely in your browser: free, unlimited, and the file never leaves your device. «Accurate» compresses the audio in your browser and transcribes it on our servers with the large Whisper model, which is far more accurate for Arabic, dialects and busy recordings. It costs 1 credit per started minute, and the first 5 minutes every day are free for your account.
Why use it
- A free, unlimited mode that runs on your device — the file is never uploaded.
- An accurate mode with the large Whisper model for Modern Standard Arabic and Gulf, Egyptian and Levantine dialects, English and more.
- Text with timings plus SRT, VTT, Word and TXT downloads in one click.
- Translate into 16 languages keeping subtitle timings, or summarize into bullets and action items.
- A clear price before you run: 1 credit per minute, 5 free minutes a day, and automatic refunds if a run fails.
- We delete your file from our servers as soon as it is transcribed, and never use it to train any model.
How to use
Choose the file
Drag an audio or video file into the box or click to choose it (MP3, WAV, M4A, MP4, MOV, WebM and more), up to 300 MB and 30 minutes. The duration and the price appear right away.
Pick the language and the mode
Choose the spoken language or leave «Detect automatically». Then choose «Accurate» for the best result (needs sign-in) or «Fast» for free transcription on your device.
Transcribe
Press «Transcribe». In accurate mode we compress the audio in your browser to a few megabytes, upload it, and show you the progress until it is done.
Download, translate or summarize
Copy the text or download TXT, SRT, VTT or Word. Below, pick a language and press «Translate» for a translation with the same timings, or «Summarize» for short bullet points.
Examples
A 25-minute university lecture
Transcribed in accurate mode for 25 credits (20 if you still have today's free minutes), and summarized into bullets for 1 more credit.
A YouTube video that needs English subtitles
An 8-minute Arabic video: accurate transcription plus an English translation (about 1,100 words = 3 credits), then download the English SRT and upload it to YouTube.
A long voice note
A 3-minute voice note in Kuwaiti dialect becomes written text for free within the daily free minutes.
| 🐇 Fast | 🦉 Accurate | |
|---|---|---|
| Where it runs | In your browser | On our servers |
| Price | Free, unlimited | 1 credit per minute, 5 free minutes a day |
| Accuracy | Good for clear speech | High, especially for Arabic, dialects and busy recordings |
| Speed | Depends on your device (faster on a computer) | Steady, with live progress |
| Privacy | The file never leaves your device | Deleted from our servers right after |
Tips
- Record somewhere quiet with the microphone close: audio quality matters most for accuracy.
- Set the spoken language instead of «Detect automatically» when the whole recording is in one language.
- Cut files longer than 30 minutes into parts with our free video trimmer (/video-trimmer/).
- Always check names and numbers before publishing; they are what any transcription model gets wrong most.
- To clean a noisy recording first, try our free audio cleaner (/audio-cleaner/).
In fast mode the file never leaves your device: the audio is read and the text is written inside your browser. In accurate mode we extract and compress the audio in your browser, then only the compressed audio (never the picture of a video) is sent over an encrypted connection to our own servers and transcribed there by a model we run ourselves (it does not pass through any AI company). The file is deleted as soon as it is done. The resulting text stays with us for 24 hours only so you can download it, then it is deleted automatically. We never use your files or texts to train any model. Translation and summaries send only the text to the AI provider for processing, and we do not keep them.
Frequently asked questions
Is it free?
Fast mode is free and unlimited because it runs on your device. Accurate mode runs on our servers: you get 5 free minutes every day, then 1 credit per started minute.
How much is a credit?
One credit is $0.05. Buy from $5 (100 credits) or bigger packs with bonus credits: $10 = 220, $15 = 345, $25 = 600. No subscription, no automatic renewal.
How much does an hour cost?
The limit is 30 minutes per file, so an hour is two files for 60 credits (about $3), minus that day's free minutes.
How are minutes counted?
Every started minute counts: 2 minutes 10 seconds = 3 minutes. We ignore the first 1.5 seconds over, which compression software adds, so a 60-second recording is 1 minute.
What if transcription fails?
If accurate transcription fails for any reason on our side, your credits are refunded in full automatically and the refund appears in your history.
Does it understand dialects?
Accurate mode handles Modern Standard Arabic and Gulf, Egyptian, Levantine and Maghrebi dialects well in clear speech. Rare slang may be written in the nearest form, so review the text.
Which formats are supported?
Most audio and video formats: MP3, WAV, M4A, AAC, OGG, OPUS, FLAC, MP4, MOV, WebM and MKV, up to 300 MB and 30 minutes per file.
Can I upload the SRT to YouTube?
Yes. Download the SRT (or the translated one) and upload it in YouTube Studio under Subtitles; the text appears with the right timings.
How does translation work?
We send the text numbered line by line to an AI model that translates each line with its number, then rebuild the SRT with the original timings. It costs 1 credit per 500 words, up to 5,000 words per run.
Do you keep my recordings?
No. In fast mode the file is never uploaded; in accurate mode we delete the audio as soon as it is transcribed and the text after 24 hours, and never use any of it for training.
Why is fast mode slow on my phone?
The model runs on your device's processor. On a computer (especially with a WebGPU graphics card) it is much faster. For long files on a phone, use the accurate mode.
When do credits expire?
Purchased credits are valid for one year from purchase, trial credits for 30 days. Questions: support@adwaty.net.
Last updated: