Start free

Talebesked til tekst.iPhone, Android, enhver app.

Telefonens optager gemmer noter som M4A, MP3 eller WAV afhængigt af enheden. Del filen til en browserfane, smid den ind, og få en tidsstemplet transskription på cirka 90 sekunder.

Drop a file, or pick one

MP3 · WAV · M4A · MP4 · MOV · MKV · OGG · OPUS · FLAC · WEBM — up to 100 MB anonymously

Paste a link, we’ll fetch the audio

YouTube · TikTok · Vimeo · Twitter · SoundCloud · Spotify · 50+ more

Record straight from your browser

Sign up takes 30 seconds — recording opens right after, in the dashboard.

No card required~90s per 60-min fileSRT · VTT · DOCX · TXTFiles auto-deleted in 24h

↓ Watch what happens

Audio in. Transcript out.

Drop a file, paste a URL, or record live — text appears back with speaker labels and timestamps. Same pipeline whether the source is a file, a URL, or your microphone.

Microphone · liveREC 00:07.41
en-US auto-detected16 kHz mono
~90s
Transcript · streaming2 speakers · 47:08
S1

Thanks for making the time. I want to start with framing — what was the original hypothesis behind the project?

S2

Honestly, maybe forty percent. The shape held — the mechanics underneath had to be rebuilt almost entirely.

S1

What flipped it for you? Was there one customer call, one piece of data

~95% accuracy on clean audioSRT · DOCX · TXT · JSON

↓ This is the dashboard

This is what loads when the job finishes.

Same layout as the real dashboard — Summary, full Transcript, Speakers tab, Exports. Key points and action items extracted automatically. Auto-tags on every job.

Try it on your own file — it's free

Three real options · honest comparison

Built-in dictation, AI speech-to-text, or a human transcriber.

Three legitimate ways to get text from spoken audio in 2026. Each is best for different work. Honest numbers below — no claim that AI matches a professional human transcriber on hard audio.

Option 01

Built-in dictation

Live, free, on-device. Handy for short bursts while you talk.

Accuracy · clear English~85%
Speaker separationNo
TimestampsNo
Languages~30
60-min fileLive only
CostFree
Best forShort voice notes while you walk. Hands-free messaging. Dictating an email at a stoplight.
Option 02

AI speech-to-text

~30× faster than realtime. 100+ languages. Speaker labels. The sweet spot for most work.

Accuracy · clear English95%+
Speaker separationYes (Pro+)
TimestampsPer sentence
Languages100+ auto
60-min file~90 s
Cost · per min$0.03
Best forInterview recordings · meeting notes · podcast transcripts · YouTube subtitles · batch jobs · API automation · field journalism.
Option 03

Human transcriber

Gold-standard accuracy when the audio is hard or the stakes are legal.

Accuracy · clear English98–99%
Speaker separationManual
TimestampsPer turn
LanguagesPer transcriber
60-min file4–8 hours
Cost · per min$1–3
Best forLegal depositions · medical dictation · archival oral history · anything where 95%+ accuracy is non-negotiable.

Built-in dictation figures from public iOS / Android speech API benchmarks. Human transcriber rates from US/UK industry surveys 2024–2025.

Accuracy · real-world numbers

95%+ on clear English. It holds up on real-world recordings too.

Modern transcription reaches 95%+ word accuracy on clear English at 128 kbps and above, comparable to a human transcriber on the same recording. The audio coming in sets the ceiling — cleaner source, cleaner transcript. The breakdown below covers the recordings we actually see in production.

97%+
Clean studio audio

USB or studio microphone in a treated room. Single speaker at conversational distance. The headline number.

95%+
Clear English at 128 kbps+

Podcast masters, interview recordings, well-mic'd meetings. The sweet spot for most professional work.

93%
Real-world podcast

Field-recorded interviews, podcast episodes at 64–128 kbps, multi-speaker recordings. Usable for editorial without a review pass.

91%
Meeting room recording

Ceiling mic, omnidirectional capture, mild reverb, multiple speakers at distance. Plan a rename pass on the speaker chips.

Common questions

7 things people ask about this.

01Hvilket filformat gemmer telefonens optager i?+
iPhone Voice Memos: M4A (AAC). Android (Google Pixel, Samsung): normalt M4A eller AAC, nogle gange MP3. Windows Voice Recorder / Sound Recorder: M4A. Mac Voice Memos: M4A. Alle disse uploades direkte uden konvertering.
02Hvordan får jeg talebeskeden fra min telefon til en browser?+
iPhone: Åbn Voice Memos, tryk på optagelsen, tryk på Del, og AirDrop til Mac eller gem til iCloud Filer. Android: Tryk på Del eller Eksporter i optageappen og send via e-mail eller skylager. Mac/Windows: filen er i en kendt mappe — træk derfra.
03Hvad er den maksimale varighed for en talebesked?+
30 minutter på Gratis. Op til 10 timer pr. fil på Pro og Business. De fleste telefon-talebeskeder er under 30 minutter, så Gratis dækker størstedelen af personlige brugssager. Forskningsinterviews og lange forelæsninger, der overstiger 30 minutter, kræver et betalt abonnement.
04Er talemærker tilgængelige for to-personers talebeskeder?+
Ja, på Pro- og Business-abonnementer. Placer telefonen lige langt mellem to talere for den reneste separation. Systemet identificerer og mærker hver stemme — klik på et mærke for at omdøbe fra 'Taler 1' til den faktiske persons navn.
05Kan den transskribere en talebesked på et andet sprog end engelsk?+
Ja — 100+ sprog, automatisk detekteret. Fransk, spansk, tysk, japansk, portugisisk, russisk og 95+ andre fungerer. For meget korte beskeder, hvor detektion er usikker, kan du tvinge sproget i de avancerede indstillinger.
06Kan jeg transskribere talebeskeder fra tredjepartsapps som Otter eller Rev?+
Hvis appen lader dig eksportere den originale lyd som M4A, MP3 eller WAV, ja — upload den fil. Vi forbinder ikke direkte til tredjepartsoptageapps. Eksporter lyden fra appen først, derefter upload.
07Hvad med talebeskeder med baggrundsmusik eller omgivelseslyd?+
Let baggrundslyd (caféstemning, stille trafik) reducerer nøjagtigheden til omkring 85–88%, men er stadig transskriberbar. Kraftig musik med samme lydstyrke som talen er den svære case — nøjagtigheden falder markant på disse sektioner.

Drop something in. See what comes out.

30 gratis minutter om måneden, intet kort. Del optagelsen fra din telefon, smid den ind — transskription på cirka 90 sekunder.

Start gratis transskription