Start free

Rogorogo MP3 ni vosa.Vakalevu talai, 100+ yabaki.

Veitikitiki MP3 file na gauna talai 64 e 320 kbps. Vakarau kena vosa talai ena gauna, vakalevu talai, 99 yabaki — walang soli file, walang re-encoding, walang tabu i queue.

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

↓ Vakacava kena vakarau

MP3 e mai. Diarized vosa talai e lako.

Keimami sa likiliki na MP3 frame headers talai — VBR, CBR, joint-stereo, kena coder ni gauna (LAME, Fraunhofer, FFmpeg). Kena file e dina na stereo ena vakalevu talai, keimami sa likitaka kena vakaloloma. Mono mix-down e vakaloloma e likiliki na pukana diarization.

interview-tape-04.mp3REC 192 kbps · stereo · 38:42
auto-detected en-GB44.1 kHz · LAME 3.100
~90s
Vosa talai · streaming95% maravi
S1

So when did you first realise the archive was incomplete?

S2

Probably around 2019, when we started digitising the reel-to-reels.

S1

And the missing tapes — were they catalogued anywhere at all?

S2

There's a paper index from '78, but half of it's water-damaged.

95% on 192 kbps stereoSRT · DOCX · TXT · JSON · VTT

↓ 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

Tolu talai dina · baleta vakasama

Vakasama Whisper. Otter o Sonix. O kami.

O iko sa dua na tekivu na likiliki Whisper ena iko na laptop vakasama kena ma computer science. Otter and Sonix e sa vinakata MP3 e veitikitiki ena subscription dashboards. Keimami e sa taucoko kena file, keimami e sa vakarau kena vosa talai, kena walang vinakataki iko e nonouti ena interface.

Option 01

Whisper vakasama / vakavulavula

Vakasama kena iko ena ma GPU kena dua ni gauna. Walang speaker diarization e vakatalia kena.

SetupPython + CUDA + 10 GB models
Speaker diarizationWalang vinakataki (pyannote add-on)
Gauna · 1 hr MP35–40 min ena consumer GPU
Yabaki99, ka lilika na model e vakaloloma e liliki na 80%
VakarauTXT / SRT / VTT / JSON
MoniVakasama + iko na kulakulanisiga
Best forTekivu na likiliki na e taucoko na GPU, walang vinaka kena speaker labels, kena vinaka na privacy kena vakasama.
Option 02

Transcription.Solutions

Veitikitiki na MP3. Vakarau kena speaker-labeled vosa talai e gauna tilaka × 0.025.

SetupVeitikitiki kina, walang account e vinaka kena supu
Speaker diarizationE solia kina (Pro & Business plans)
Gauna · 1 hr MP3~90 seconds
Yabaki99, auto-detected
VakarauSRT · VTT · DOCX · TXT · JSON
Moni · per min$0.03
Best forDua ni keda na e taucoko na MP3 — journalist tape, podcast export, voice memo, archival dub — na e vinaka talai accurate na vosa talai.
Option 03

Otter / Sonix

Vakasolo dashboard, monthly minutes cap, English-tuned. File upload e vakamacala na feature.

SetupAccount + moni plan
Speaker diarizationAcoustic, EN-leaning
Gauna · 1 hr MP35–10 min ena queue
YabakiOtter EN-only; Sonix ~40
VakarauSaasoma kena moni tiers
Moni$17+/mo o $10+/hr (Sonix)
Best forTeams na e vinaka kena transcript editor kena collaboration UI na kena sa liliki clean na API-style file→vosa flow.

Pricing and feature availability accurate as of May 2026. Whisper performance varies by model size and hardware.

Solia kena MP3

Tolu talai na e saqa na keda ena generic transcription tools.

MP3 e dua na format, walang rogorogo style — kena marau kena failure modes e mai kena encoder, walang kena vosa.

Kena saqa talai

  1. 1VBR headers e walang sa likiliki dina. Vakalikilika tools e sa likiliki variable-bitrate MP3s ena fixed-rate kena miscalculate duration — timestamps e vakasikasika kena minutes ena hour-long file.
  2. 2Joint-stereo e sa vakaloroko ni mono durante e upload preprocessing. O iko e vakaloloma kena per-speaker channel separation na e dina talai kena file.
  3. 3Embedded ID3 album art e saqa vakalikilika uploaders — e sa vakaloloma na file ena 'walang pure audio' o e sa vakaroroko, kena gauna kaukauwa quality.

Kena keimami e likitaka

  1. 1Keimami e sa vakarau na Xing/LAME header kena vinakata kena frame-count fallback kena liliki. VBR timestamps e vakasolo kena maravi ±0.1 s ena multi-hour files.
  2. 2Joint-stereo kena true-stereo MP3s e decoded ni L/R PCM kena diarization. Kena iko na speakers e panned, keimami e vakasolo kena split.
  3. 3ID3v1, ID3v2, APE tags, embedded art — e sa vakaloloma talai. Keimami e walang re-encode kena iko na MP3.

Vinaka na job settings na MP3 uploads

Defaults na e vinakata ~80% na MP3 files. Override per-job mai kena form.

Decoder
Frame-accurate, walang re-encode
Diarization
Channel split kena stereo, liliki acoustic
Speaker model
Auto · 1-12 speakers
Yabaki
Auto-detect mai kena first 30 s
Filler words
Vakaroroko (toggle kena vinaka)
Vakarau bundle
DOCX + SRT + timestamped TXT

Accuracy · real-world numbers

95%+ ena 192 kbps stereo. Dina e liliki ena 64 kbps mono.

MP3 maravi e solia kena kena encoder vinaka, walang keimami. Perceptual compression ena ~96 kbps e sa vinaka talai na rogorogo pukana dina; e liliki na 64 kbps, sibilants kena consonants e sa vakarorogo. Numbers e mai na real customer MP3s ena production.

96%
320 kbps stereo, studio source

Kena sa liliki walang lose na pukana kena vosa. Podcast masters, dictation app exports, professional interview rigs. Diarization e maravi kena speakers ena vakalevu talai channels.

95%
192 kbps stereo, 2-3 speakers

Kena gauna kaikaikavi talai bitrate na vosa-word MP3s. Zoom exports, Riverside downloads, voice recorders default. Compression artifacts e walang marivarivaki kena recognizer.

91%
128 kbps mono, conversational

Voice memo defaults ena kaikaikavi phones. Acoustic diarization e likitaka 2-4 speakers. Numbers kena proper nouns occasionally e vinaka na likiliki.

84%
64 kbps mono, archival / phone-dump

Kena answering-machine rips, lecture archives, narrow-band sources. High-frequency consonants (f/s/sh) blur. Kena liliki dina talai — plan a proofread.

Vakakavi ni tu

8 talai na keda e vakakavi baleta MP3 transcription.

01Kena lilika talai MP3 bitrate na e vinakata na transcript?+
64 kbps e kena lilika talai liliki. E liliki na, sibilants (s, sh, f) e sa vakaroroko kena noise kena word error rate e liliki na 20%. Kena iko e lilikitaki gauna vou, target 128 kbps mono o 192 kbps stereo — dua talai e liliki kaikaikavi kena vosa.
02E vinaka na soli kena iko MP3 ni WAV kena kalougata?+
Walang. Re-encoding MP3 → WAV e liliki zero maravi kena kena data na encoder e vakaloloma e liliki talai. Veitikitiki na MP3 dina kena. Keimami e sa decode frames kena memory kena vakarai PCM kena recognizer.
03Kena stereo MP3 e vinakata talai na speaker labels na mono?+
Talea kena speakers e dina talai recorded ena vakalevu talai channels — kaikaikavi stereo MP3s e kena gauna talai ena vakalevu talai sides ('dual mono') kena gauna liliki. Dina channel-split (e.g. Riverside exports, two-mic field rigs) e vinaka kena keimami vakasikasika acoustic diarization kena talai speakers near-perfectly.
04Kena kaikaikavi talai MP3 file size na o iko e vinaka?+
5 GB per upload, kena liliki ~60 hours ena 192 kbps o 90 hours ena 128 kbps. Kena file na e kaikaikavi keimami e liliki chunked upload — walang vinaka na split kena iko.
05Kena 60-minute MP3 e gauna talai kena rogorogo?+
Kaikaikavi 90 seconds mai upload-complete ni transcript-ready, walang bitrate. Decoding MP3 frames e gauna kaukauwa; kena gauna e kena recognizer. Diarization e liliki 5-10 seconds ena multi-speaker files.
06Kena iko na MP3 e taucoko na background music — kena vosa talai e saqa ni?+
Quiet bed music under vosa e maravi talai. Loud music na e wilika kena voice (intro stings, scoring under interviews) e liliki misrecognitions ena overlapping syllables. Toggle music suppression kena job form kena pre-filter.
07O iko e likitaka na MP3s ripped mai phone voicemail o answering machines?+
Io, kena liliki vakalikilika e 8 kHz narrow-band re-encoded ena MP3 — kena pukana maravi ceiling e solia kena original PSTN capture, walang MP3 wrapper. Expect 78-85% maravi kena lilika na source, kena liliki keimami e vakarau kena underlying call.
08O iko e vakasolo kena iko na MP3 e vosolia na kena vosa talai?+
Files e vakaroroko e liliki 30 days kena default, o talai kena request ena dashboard. Kena vosa talai e vakasolo kena iko na account kena liliki iko sa vakaroroko. Keimami e walang vakarau customer audio kena likiliki dua na model — never.

Veitikitiki na iko na MP3. Vakarau na vosa talai ena 90 seconds.

30 vakasama minutes ena gauna kaikaikavi. Walang kadi vinaka. Speaker labels, 99 yabaki, kaikaikavi vakarau formats vinakataki.

Qai vakasama