Audio and video transcription online
Upload an interview, call, podcast or video to get a transcript with timestamps. Enable speaker detection and audio events if needed. You will see the character cost before transcription starts.
Transcribe fileSign in to transcribe a file. You can use your ruble balance, bonus rubles, or remaining legacy characters. View top-up amounts.
Audio and video
Upload audio or a video with sound. MP3, WAV, MP4 and other formats listed below are supported.
Multiple speakers
Enable speaker detection and specify the number of speakers. The transcript will separate their lines.
Timestamps and events
Timestamps appear next to each line. Audio event labels can be enabled separately.
What the result looks like
Demo: fictional dialogue illustrates timestamps, speakers and audio events.
How do you usually start working on a new project?
With the task. First we understand what the person wants to do and for whom.
Short pause
When can you tell that a solution really works?
When people can use it comfortably without extra explanation.
Use cases
Turn your recording into text for further work.
How to turn audio or video into text
- Sign in, check your balance and select a file to transcribe.
- Wait for the recording to be checked. The form will show its duration and processing cost.
- Choose a language or auto-detection. Optionally enable speaker separation and select between 1 and 9 speakers.
- Start transcription, check the text against the recording and copy the result.
Before publishing, check names, numbers and technical terms against the recording. Background noise and overlapping speech can cause transcription errors.
File formats and limits
Maximum file size — 512 MB. Supported extensions: AAC, FLAC, M4A, MKV, MOV, MP3, MP4, MPEG, MPGA, OGA, OGG, OPUS, WAV, WEBM.
Select Russian, English, German, Spanish, French or auto-detection. Video files must contain an audio track.
Wait for the current task to finish before starting another one.
How transcription pricing works
Legacy characters are used first. Their previous billing method still rounds duration up to whole minutes. If you do not have enough characters, the remaining portion is paid in rubles. Before starting, the form shows how many characters and rubles will be charged.
Cost in rubles = recording length in seconds ÷ 60 × price per minute. Duration is not rounded up to whole minutes: 30 seconds cost half the per-minute rate, and 90 seconds cost one and a half times that rate. The current rate and total are shown after the file check, before transcription starts.
Topping up on the pricing page adds funds to a shared ruble balance for text-to-speech and transcription. Transcription has its own rate per minute of recording; text-to-speech prices per 1,000 characters do not apply.
Questions about transcription
Can I try transcription using bonus rubles?
Yes. After signing up, open the balance page, click “Get 50 ₽” and follow the Telegram bot’s instructions. The one-time bonus is credited to your shared balance and can be used for text-to-speech and transcription.
What happens to my balance during processing?
The job cost is reserved when processing starts. After successful transcription, the reserved amount is charged. If processing fails, the reserved rubles and characters are returned to your balance.
How do I save the transcript?
Click Copy text in the result and paste the transcript into a document or text editor.
What are speakers and timestamps for?
Timestamps help locate sections in the original recording. Speaker separation groups dialogue when enabled.
Can I use the transcript for an article or subtitles?
Yes. Copy the text into an article or notes. Use the transcript and timestamps in your video editor to create subtitles.