TAK! TEXT
Telegram bot Transcription · translation · AI

Voice
to text,
in seconds

Send or forward a voice message, video note, file, or link — the bot returns text with timestamps and speaker labels. Up to 48 speakers, 90+ languages, files up to 2 GB.

Works directly in Telegram, no signup required. Free minutes every month for everyone. Paid plans for those who receive dozens of voice messages a day.

TAK! TEXT online
/start 21:31 ✓✓
🎤 TAK! TEXT — audio & video to text in seconds

Send a voice message, audio, or video — and get text back.

90+ languages · translation · summary · Q&A

21:31
Bot info
0:38 ● 148 KB
21:32 ✓✓
Transcribing audio.
📎 Voice message · 0:38
Transcription ready:
copy

[Speaker 1]: [00:00] Should we move the meeting from Thursday to Friday?

[Speaker 2]: [00:14] Sure, Friday at three. I'll send a calendar invite.

[Speaker 1]: [00:27] Perfect, thanks!

21:32
📥 Download 📝 Summary
🌍 Translate 💬 Ask AI
Transcription settings
10:23 ● 3.2 MB
21:35 ✓✓
Transcribing audio.
📎 Voice message · 10:23
90+languages
/
48speakers
/
2 GBfile
/
20+platforms
/
24/7in Telegram

Six things
the bot does

Details and nuances — in FAQ →

01.

Two recognition modes

"Speed" delivers results in seconds. "Quality" is more accurate but may be a bit slower. Perfect for important recordings. Switch with one tap — both available on all plans.

02.

Speakers and timestamps

Speaker diarization — up to 48 speakers in a single recording, each labeled. Timestamps every ~30 seconds for navigating long recordings.

03.

90+ languages

Automatic language detection. You can also select the language manually — this significantly improves accuracy for noisy recordings and specific dialects or accents.

04.

Files up to 2 GB and links

Lectures, podcasts, hours-long meetings — in a single file. For many platforms you can just send a link: YouTube, Vimeo, SoundCloud, Dropbox, Twitch, Dailymotion, and 15+ more.

05.

AI tools for the transcript

Summaries of long recordings. Ask a question about the transcript and get an instant answer. Translate the transcription into other languages.

06.

Data privacy

Audio is deleted immediately after transcription. Transcripts are deleted after 24 hours. No sensitive user data is stored. Servers in Germany (Hetzner) — GDPR-compliant. More about security →

Why professionals forward voice notes to us

Journalists & podcasters Interviews with multiple speakers, dictated notes, press conferences. Speakers are labeled, timestamps help find the right quote.
Students & researchers Two-hour lectures into notes. AI summaries for review, timestamp navigation. Ask a question about the transcript.
Business Voice messages from managers and colleagues into text. Forward a long audio from a group chat — get a summary and a list of action items. Meeting minutes with clear speaker separation.
Content creators Podcasts into show notes, videos into subtitles, translating clips for international audiences.

You only pay for minutes

Monthly subscription or minute packs with no commitment. No limit on the number of recordings. Free minutes every month.

FREE $0€0/mo START $3.49€3.19/mo PRO $10.79€9.69/mo POWER $32.49€29.49/mo
Minutes per month 30 first month
then 15/mo
300 3,000
Single recording length 5 min 30 min unlimited
File size 300 MB 2 GB 2 GB
AI requests / mo 10 50 Unlimited
Transcription by URL
Speaker separation
Timestamps
AI summary
Export PDF / TXT
Start Subscribe Subscribe
FREE $0€0/mo
  • 30 min first month, then 15 min each month
  • Files up to 5 min · 300 MB
  • 10 AI requests
  • Speakers, timestamps, export
Start
START $3.49€3.19/mo
  • 300 min/mo (5 hours)
  • Files up to 30 min · 2 GB
  • 50 AI requests
Subscribe
POWER $32.49€29.49/mo
  • 3,000 min/mo (50 hours)
  • Unlimited file length · 2 GB
  • Unlimited AI requests
  • For teams and frequent use
Subscribe

Prefer not to subscribe? Minute packs from $1.29€1.19 — 70 to 2,000 minutes, stack as many as you need. START plan limits apply (see table above).

Final amount may vary slightly with current exchange rates and payment processor fees.

Frequently
asked questions

Six questions we get most often. Full list — on the FAQ page →

Q1.

What file formats are supported?

All popular audio and video formats: MP3, WAV, OGG, FLAC, M4A, MP4, MOV, WebM, and more. Also voice messages, video notes, and links from 20+ platforms.

Q2.

How accurate is the transcription?

On clean speech — above 95%. The bot selects the best model and parameters for the specific language, recording type, and audio quality.

Q3.

What happens to my files?

Audio and video files are deleted immediately after transcription. Transcripts are stored encrypted for up to 24 hours, then permanently deleted. Diagnostic logs are minimized and do not store transcript content. Servers — Hetzner, Germany (GDPR).

Q4.

Which platforms are supported for links?

YouTube, Vimeo, Dailymotion, SoundCloud, Dropbox, Twitch, Rumble, Bandcamp, Mixcloud, and more — 20+ total. Available on PRO and POWER.

Q5.

What's the difference between Speed and Quality modes?

"Speed" delivers results in seconds. "Quality" is more accurate and supports more languages but may be a bit slower. Perfect for important or noisy recordings. Both modes are available on all plans and switch with one tap.

Q6.

Are AI tools free?

Summaries (AI) are free on all plans, including FREE. Questions and translation use AI requests: 10 on FREE, 50 on START, unlimited on PRO and POWER.

Stop
replaying
voice messages