Transcribe · Live subtitles · Text to speech
Transcribe audio & video.
Pay as you go.
Upload audio or video, get a clean transcript, subtitles, summary and translation in minutes. 99 languages, up to 97% accurate. No subscription - credits never expire.
No card needed to start 15 free minutes when you sign up
Up to 97% accurate
99 languages
No subscription
Pay per use
Starting at $0.83/hour
Audio deleted after processing
The math
Otter, TurboScribe & co. charge $10–$20/month - whether you use them or not.
Speecho is $0.83–$1.25 per hour of audio - and credits never expire.
One 90-minute interview a month? You’d pay about $1.50 instead of $20.
Features
Everything you need
Fast
Transcription in minutes, not hours. Most short files done in under a minute.
Accurate
Up to 97% word accuracy.
99 languages
Auto-detect or specify the language.
Audio + video
MP3, M4A, WAV, MP4, MOV, WEBM. Video audio extracted in your browser.
Text to speech
Turn any text into natural speech - 512 voices, MP3 download.
No subscription
Top up when you need it. Credits never expire.
Edit in place
Fix names and terms right in the transcript - every download and subtitle updates instantly.
Listen while you read
Optionally keep the audio: click any timestamp to replay that moment. Off by default, deletable anytime.
Video captions
Burn subtitles into Reels, TikTok and Shorts, with word-by-word highlight, previewed on your own frame.
How it works
How it works
No setup required. Start transcribing in seconds.
Watch the full tutorial on YouTube
1
Upload your file
Drag and drop any audio or video file.
2
AI transcribes it
Our AI processes your file in seconds.
3
Get the text
Copy the result or receive it by email.
Outputs
One upload, four outputs
Everything you need from your audio or video in a single transcription.
Clean paragraphs with natural breaks. Optional timestamps. Download as text, Markdown, Word or PDF.
00:12So today we're talking about how we ship faster without burning out the team.
00:38I think the key insight is async-first communication - let people focus.
01:04Right, and pairing only when it genuinely unblocks…
Caption files ready for video editors, Premiere, Final Cut and YouTube.
SRTVTT
1
00:00:12,400 → 00:00:15,800
So today we're talking about
how we ship faster. Recap with action items - or pick a style: meeting minutes, interview notes, lecture notes, podcast show notes.
Key points
Async-first comms ship faster than meetings
Pair only when truly blocked
Action items
Draft async-first guideMaya
Translate the full transcript to 18 languages. Cached - re-open instantly.
EN → Cached
ES FR DE PT IT JA ZH KO +10
Speaker identification Optional 5th layer
Auto-label each voice in interviews and panel recordings - Speaker 1, Speaker 2, and so on.
Also in Speecho
Same credits, more tools
Everything here runs on the balance you already have.
Live subtitles Captions on screen as you speak, translated live. Meetings, lectures, OBS streams. Learn more → Live interpreter The translation spoken aloud in a natural voice. Listeners join by QR code on their own phones. Learn more → Conversation translator One phone between two people. Each reads the other in their own language, and the talk is saved in both. Learn more → Text to speech 512 voices, free previews, MP3 download. Dialogues with a different voice for every character. Learn more → Teleprompter Free, in the browser. The script scrolls as you speak and waits when you pause. Learn more →
Pricing
Simple pricing
Top up once, use anytime. The more you add, the more you save.
Basic
$5
2,400 credits
- 4 hours transcription
- or
- up to 48 min text to speech
- or
- up to 1 hour live interpreter
- $1.25 per hour of transcription
- All file formats · 99 languages
- AI summary & chapters
- Translation to 18 languages
Recommended
Pro
$10
6,000 credits
- 10 hours transcription
- or
- up to 2 hours text to speech
- or
- up to 2.5 hours live interpreter
- $1.00 per hour of transcription
- All file formats · 99 languages
- AI summary & chapters
- Translation to 18 languages
Team
$25
+1,800 bonus credits 15,000 credits
- 28 hours transcription
- or
- up to 5.5 hours text to speech
- or
- up to 7 hours live interpreter
- $0.89 per hour of transcription
- All file formats · 99 languages
- AI summary & chapters
- Translation to 18 languages
Best value
Studio
$50
+6,000 bonus credits 30,000 credits
- 60 hours transcription
- or
- up to 12 hours text to speech
- or
- up to 15 hours live interpreter
- $0.83 per hour of transcription
- All file formats · 99 languages
- AI summary & chapters
- Translation to 18 languages
Prices in USD. VAT may apply at checkout based on your country.
Yes. Create an account and confirm your email - you get 15 free minutes of transcription, no credit card required. Welcome credits are regular credits, so you can also spend them on handwriting scans or video captions. After that it is pay as you go: top up from $5, credits never expire.
No. Your credits never expire - top up once and use whenever you need.
Speecho supports 99 languages including English, Spanish, French, German, Chinese, Japanese and more.
Yes, a free one: sign up with your email or with Google. Confirming your email adds 15 free minutes, and you can look around the app with sample data before you sign up.
Audio: MP3, M4A, WAV, WEBM up to 200MB. Video: MP4, MOV, WEBM up to 4 hours - we extract the audio in your browser before upload, so even large recordings transfer fast.
We use modern AI speech recognition: 90–97% word accuracy on clean audio, in 99 languages.
Yes. Every transcription includes .srt and .vtt subtitle files - ready for Premiere, Final Cut, DaVinci Resolve or YouTube. Timing is synced to the audio.
Yes. Click "AI summary" on any finished transcript to get key points and action items - or pick a template that matches your recording: meeting minutes, interview notes, lecture study notes or podcast show notes. The first two note styles per transcription are free; additional styles cost a few credits.
Yes. Every transcript downloads as plain text, a formatted Word document (.docx) or a PDF with speaker labels - alongside .srt/.vtt subtitle files.
Yes. One click turns the transcript into a ready-to-send follow-up email - a short recap, the agreements and next steps. Copy it or open it straight in your mail app.
Yes. Translate any transcript to 18 languages with one click. Translations are cached, so re-opening them is instant. The first two languages per transcription are free; additional languages cost a few credits.
When enabled before upload, Speecho labels each voice (Speaker 1, Speaker 2, …) - useful for interviews, podcasts and meetings. It costs nothing extra: the price is the same as a standard transcription.
Yes. Open any transcript and click "Ask AI" to chat with it - ask what was decided, what a speaker said or where a topic came up. Answers come strictly from the transcript. Each question costs a few credits, priced by the transcript length (from 5 credits).
Yes. One click splits the recording into chapters with clickable timestamps - jump straight to the part you need. Chapters are free and stay saved with the transcript.
Yes. Click the pencil next to any paragraph, fix names or terms and save - every download format (text, Markdown, Word, PDF, subtitles) uses the corrected text. Speaker labels can be renamed the same way.
Yes - tick “Keep audio” before uploading (it is off by default). The recording is stored privately, a player appears above the transcript and clicking any timestamp jumps to that moment. You can delete the audio at any time; the transcript stays.
Yes. Pick the “Clean text” style in the AI notes menu to get your transcript without the ums, false starts and repeated words - nothing is summarized or reworded, only the noise is removed.
Add them to the glossary (Names and terms) before uploading - product names, people, industry jargon. Speecho keeps the glossary and applies it to every future transcription automatically. The same glossary works in live subtitles, where you can also set how a term is translated.
Yes. Transcribe each file, then select them in your history and press Merge: one transcript with the parts in the order you set, timestamps running on and speakers numbered across the parts. Up to 20 at once, free, the originals stay, and if every part kept its audio, the audio is joined too.