Speak the way you always do — Iskrio turns messy speech into clean text. In any app, wherever the cursor is.
Free forever — up to 1,500 words a month. No card.
For developers · writers · managers — a format for each
You speak naturally. The text comes out right.
One raw stream of speech — three different results. Switch the mode and see.
so um we need to move the contractor meeting to thursday… no wait friday, friday around three and uh remind them to send the estimate I mean the final version of the estimate
We need to move the contractor meeting to Friday at 3:00 PM. Remind them to send the final version of the estimate.
Please reschedule the contractor meeting to Friday at 3:00 PM. Please also remind the contractor to submit the final version of the estimate.
• Contractor meeting → Friday, 3:00 PM
• Reminder: send the final version of the estimate
Works where you write
…and any other app
Speak faster than you type
Typing runs at about 40 wpm, speech runs 120–160. You get the same text 3–4× faster. Iskrio strips away everything extra:
Built for people who work with text all day
Developers & vibe-coders
Prompts, code comments and docs — by voice, in any IDE. It gets technical terms and mixed Russian-English speech right.
Writers & copywriters
Articles, posts and scripts by voice — 3–4× faster than typing. Your train of thought keeps up while you talk.
Managers & leads
Emails, reports and minutes — by voice. Autoreplace fills in boilerplate phrases for you.
Everything to write with your voice — anywhere
Iskrio works in any app and turns your speech into clean text — no manual editing.
In any app
Browser, email, messengers, notes, code. Dictate where you normally type — no switching.
AI polishes the text
Removes "umm", repeats and slips, adds punctuation and formatting — clean writing comes out.
Understands plain speech — and tech
Dictate emails, notes and messages in everyday language. And when tech words slip in, the 200+ term dictionary fixes them: "engine-x" becomes Nginx.
Context-aware
Knows where you're dictating: casual tone in messengers, business tone in email, exact terms in code editors.
Modes for every task
AI cleanup, Verbatim, Formal, Key points, Schedule or your own prompt — one hotkey, different results.
Privacy
Your text history and dictionary stay on your device only. Audio is not stored after processing.
Three steps — and your text is ready
No setup. Install it and start dictating right where you usually type.
Press a hotkey
Anywhere — in the browser, email or editor. Iskrio launches instantly and starts listening.
Just speak
Dictate naturally, like a conversation — with pauses, in Russian, English and other languages.
Get clean text
AI removes the clutter, adds punctuation and inserts the finished text right where you write.
What goes to the cloud — and what never does
The first question about any dictation tool is where your words end up. Here's the straight answer.
Stays on your device
- Your entire text history
- Personal dictionary and its auto-learning
- Settings and app context
In the cloud — only during processing
- Audio is sent for recognition and not stored
- We don't keep recordings or train models on them
- Only the app category is sent — never your window titles
Simple pricing — no surprises
Start free. Move to Pro when dictation becomes a habit.
- 10 minutes per month (up to 1,500 words)
- All AI modes, context-aware style
- Mixed speech, 90+ languages
- IT dictionary + 10 custom terms
- 120 minutes per month
- Everything in Free
- Personal dictionary: up to 50 terms
- Email support
- 300 minutes per month
- Everything in Plus
- Unlimited dictionary + auto-learning
- Priority support
- 500 minutes per month
- Everything in Pro
- Maximum minutes for heavy use
- Priority support
Common questions
The essentials, briefly. Everything else is in the docs.
How is it different from built-in Windows dictation?
Built-in dictation gives you a raw transcript — every "umm", repeats, no punctuation — which you then clean up by hand. Iskrio returns finished text: it strips the clutter, adds punctuation, formats it and styles it for the app you're in. It also handles tech terms and mixed Russian-English speech that built-in dictation garbles.
Which apps does it work in?
Any: if you can type there, you can dictate there. Telegram, Word, your browser, your code editor — Iskrio inserts text into the active field on a hotkey.
Does it work with background noise or an accent?
Usually yes — recognition copes well with background noise and with less-than-perfect diction or an accent. A lot still depends on conditions: a quiet room and a closer mic always give cleaner text. The easiest way to check is on your own voice — that's what the free plan is for.
Do I need internet?
Yes, recognition and AI processing require a connection. An offline mode is on the roadmap.
Which languages are supported?
Russian and English first, including mixed speech. Other languages are supported too, and tech terms are corrected against the built-in dictionary.
Where does my voice go?
Audio is processed in the cloud and not stored after recognition. Your text history and dictionary stay on your device only. Details are in the Data Policy.
Is it free?
Yes: the free plan gives you 10 minutes of dictation a month, forever (up to 1,500 words), no card needed — just sign up with your email. Paid plans are for when dictation becomes a daily habit.