Transcribe Russian audio to text
Transcrib Si runs Russian speech on its own transcription engine, tuned for Russian and for the Russian-English mix that real work produces, and returns text with timestamps and speaker labels. One file is up to 64 MB and 120 minutes, the free plan is $0 with 3 files a day, and Pro is $10 a month when you pay yearly.
Why Russian needs its own engine
Every general speech-to-text tool we tried handles Russian like a guest: tolerated, not understood. The words arrive, but the product terms get mangled, the numbers turn strange, and a two-hour conversation in mixed Russian and English comes back as a wall of near-misses. That is why this service exists.
Two languages, and the mix between them: Russian, and English, including speech that switches between them inside one sentence. Both stay as they were spoken.
| Case | What happens | What you get back |
|---|---|---|
| Russian speech | transcribed as Russian, with terms and numbers as they were spoken | text, timestamps, speakers |
| Russian mixed with English | both languages kept, no translation in between | text, timestamps, speakers |
| Pure English | the cloud fallback provider, when it is enabled on the server | text, timestamps, speakers |
What happens to your file
The short version: a Russian or mixed file stays on our own infrastructure, and a file that is entirely in English may be passed to a third-party cloud speech provider in the United States, but only when that fallback is enabled on the server. Full detail in the privacy policy.
| File | Where it is transcribed | What leaves your account |
|---|---|---|
| Russian or mixed Russian-English file | our own engine, on our infrastructure outside the United States | the audio and the transcript |
| Pure English file | a third-party cloud speech provider in the United States, when that fallback is enabled on the server | the audio and the transcript |
| Any file | never used to train models for other customers, never sold | — |
What we promise and what we do not
| Claim | Status |
|---|---|
| A timestamp for every segment | yes, part of the contract |
| Speaker labels | yes, on request |
| Automatic language detection | yes — 0.39 s on a short file in our own measurement |
| A single accuracy percentage | no — we do not publish one, and we will send the evaluation report and the sample set on request |
| “99% accuracy” | no — a marketing number without a test set behind it |
| More than two languages | no — Russian and English today |
Russian and English in the same sentence
Mixed speech is the normal case in Russian offices, and it is where a general model loses the thread. Our engine keeps both languages in place, which is what the example below shows — the same segment structure, the same timecodes, no translation in between.
| Time | Speaker | Text |
|---|---|---|
| 00:04 | Speaker 1 | Мы перенесли релиз на четверг, потому что QA ещё закрывает два бага. |
| 00:11 | Speaker 2 | Okay, but the API rate limit is still 60 requests per minute. |
| 00:18 | Speaker 1 | Понял. Я подниму лимит до 300 и напишу в Slack. |
The numbers
| Value | Where it comes from | |
|---|---|---|
| One file | 64 MB and 120 minutes | Service contract |
| Free plan | $0, 3 files a day, up to 30 minutes each | Pricing page |
| Pro | $10 a month paid yearly ($120 a year) or $20 a month paid monthly | Pricing page |
| Pro fair use | 1,500 minutes of audio per calendar month | Pricing page |
| Export formats | 6 — TXT, Markdown, SRT, VTT, CSV, DOCX | Pricing page |
| Languages | 2 — Russian and English, set or detected | Service contract |
Put a file on the page and read the first lines.
The widget below is the same one as on the home page: file → transcript with timecodes and speakers → export. The transcript is saved to your account.
Drop an audio or video file
or — mp3, wav, m4a, mp4, up to 64 MB and two hours
Transcription runs under your account: sign in or create a free account.
No account yet? The free plan needs no card. Create a free account or sign in.
Questions this page answers.
Do you support languages other than Russian and English?
No. Two languages today: Russian and English, including speech that switches between them inside one sentence. We would rather say that plainly than show a list of a hundred flags.
What happens to my recording?
It is used to produce the transcript and to run the service. We do not sell it and we do not train models on it for other customers. Delete a transcript in the app and the audio goes with it. A Russian or mixed file is processed on our own infrastructure, which is outside the United States; a file that is entirely in English goes to a third-party cloud speech provider in the United States when that fallback is enabled on the server. The details are in the privacy policy.
How good is the Russian?
Our engine runs a model profile tuned for Russian speech and for the Russian-English mix. We do not publish a single accuracy number on the site: ask for the evaluation report and the sample set and we will send both.
Does the service translate the transcript?
No. The transcript is in the language that was spoken. Translation is not part of the product today.
Can I dictate in Russian and get the text?
Yes — a voice memo is just a small audio file. There is a guide on transcribing voice memos.
Is there a Russian-language interface?
The interface is in English today. The service handles Russian audio, and the API, the MCP server and the plugins are language-neutral.
Related pages.
The general guide: formats, limits, exports and prices.
Как расшифровать аудио в текст
Тот же материал на русском.
Meetings in mixed Russian and English are the normal case.
Where a file is processed and how long it is kept.
Plans, limits and fair use.
Mark Burmistrov, Founder.
Every number on this page comes from the service contract, the pricing page or our own measurements. The author of the page and the publisher of the site are the same person. Questions and corrections: support@transcrib.si.