Every voice message is a productivity trap: you can't skim it, search it, quote it, or drop it into your CRM. Playing recordings back wastes hours you don't have — for coaches, consultants, agencies, and sales teams, this is a daily tax.

Forward a voice note, audio file, or video to your Telegram bot. Seconds later, it replies with clean text, Markdown, or a ready-to-share Word document (.docx) — in the language you choose. It runs entirely on YOUR n8n and YOUR Telegram bot — no $20/mo transcription subscriptions, no third-party SaaS reading your private chats.

✦ Why buy this instead of wiring 3 nodes yourself? A basic "Telegram → Whisper" flow takes 10 minutes to build and breaks on the first real user. This is a 55-node, battle-tested application with an interactive UI, state management, built-in AI translation, native .docx file generation, localization, and full error recovery already engineered for you.

✦ What you get

Lightning-fast: transcribes a 30-second voice note in ~2–5 seconds right in the chat (powered by Groq Whisper's free tier)

AI translation built in: Whisper auto-detects the spoken language, and the bot delivers the transcript in the language you pick — 7 quick-pick buttons + a full menu of 90+ languages (per-line translation, timestamps preserved)

3 output formats: plain text, Markdown, or Word (.docx) — generated natively inside n8n (no Google Docs or external APIs required)

Smart memory: remembers each user's preferred language — 1-tap transcription for repeat use

Auto-localized UI: bot buttons and messages automatically adapt to the user's language (EN, RU, DE, FR, ZH, JA)

✦ 100% Private by design

Zero middlemen: audio and transcripts travel strictly between your Telegram bot, your n8n instance, and your chosen AI engine

Full data ownership: we never see, log, or touch your data. On self-hosted n8n, your transcripts never leave your server

✦ Built to survive real users (production-ready)

20 MB size guard: catches oversized files with a friendly message before hitting Telegram's hard API limit

Foolproof inputs: replies politely to accidental photos, stickers, and random text instead of failing silently

Double-tap protection: prevents duplicate jobs and double API billing if an impatient user mashes buttons

Self-healing state: automatic state cleanup and error recovery

Instant admin alerts: optional 1-toggle setting sends you a private Telegram alert the moment a job fails (includes error log, chat ID, and timestamp)

✦ Zero-cost default + plug-and-play AI engines

Default: Groq Whisper (free tier) — handles thousands of transcriptions per day at $0 cost with blazing speed; translation runs on the same free Groq key

✦ 10–15 minute setup for non-programmers

Import 1 JSON file into n8n (cloud or self-hosted)

Paste 2 free credentials (Telegram Bot Token from @BotFather + free Groq API key)

Toggle Active — you're live. Includes a step-by-step INSTALL.md guide with screenshot-level details and a quick troubleshooting table.

✦ Requirements (please read before buying)

n8n (cloud or self-hosted), a Telegram bot token from @BotFather, a free Groq API key

Telegram registers a webhook, so your n8n must be reachable over HTTPS: n8n Cloud does this automatically; on a self-hosted server you need a domain with HTTPS (one-time setup, covered in the guide)

If your n8n runs only on localhost/LAN without a public address, use n8n Cloud — the bot needs a public URL to receive messages

✦ What's inside the delivery

transcriber-workflow.json — clean, organized 55-node n8n workflow (credentials stripped and ready for yours)

INSTALL.md — complete setup and customization manual

✦ Want to plug this into Slack, WhatsApp, or your CRM? Because the core transcription pipeline is channel-agnostic, it can easily power Slack, Discord, WhatsApp, email, or web forms.

Need custom features? I built this architecture from scratch. Message me if you want me to tailor this workflow for your business — including speaker diarization (who said what), AI meeting summaries / action items, CRM auto-logging (HubSpot/Notion/Pipedrive), or high-volume scaling.