Who said what, when.
Not a wall of words. Every voice on its own line, every line stamped against the recording.
Customer interview
38 MIN · 2 SPEAKERS · ENGLISH
Sarah
Walk me through what happened after the rollout, start to finish.
Tom
We shipped on a Friday, which was our first mistake. By Monday the support queue had doubled and the team spent the whole week digging out.
Sarah
Doubled from what baseline? I want the actual numbers before we present this.
Tom
From about forty tickets a week to just over ninety. Most of them were password resets caused by the migration, so the fix is one email.
Speakers, separated
Every voice is detected and gets its own line, so interviews and meetings stay readable.
Timestamps throughout
Every line is stamped, so any quote can be checked against the recording in seconds.
Long recordings, handled
Hours-long files are split, transcribed in parallel, and stitched back in perfect order.
Five languages
English, Dutch, German, French and Spanish, with the same accuracy in each.
The transcript is just the beginning.
Most tools stop at the words. Here, one click turns them into meeting minutes, action items, a flowchart, or the follow-up email.
Q3 kickoff — Meridian rollout
TODAY · 32 MIN · 5 SPEAKERS
Upload. Transcribe. Read.
Drop in any file
Audio or video, phone memo or studio file. Up to 1GB on the free plan.
We do the work
Speakers separated, timestamps added, big files handled automatically.
Read, search, share
A clean transcript you can search, copy, and hand to anyone.
Your recordings are your business. We keep it that way.
Encrypted, always
In transit and at rest, on every recording and transcript.
Trusted infrastructure
Built on Google Cloud and vetted AI providers, nothing home-rolled.
Hosted regionally
EU-based company, GDPR-native, regional data hosting.
Yours to delete
Your audio is never training data, and deleting a recording deletes it everywhere.
Before you upload.
How accurate is the transcription?
Very accurate on clear recordings: state-of-the-art speech models with speaker separation and timestamps on every line. Accuracy tracks audio quality, so a phone on the table works, and a good microphone works better.
Which formats and file sizes work?
M4A, MP3, WAV, MP4, MOV, WEBM, FLAC, and OGG, up to 1GB per file on the free plan and larger on paid plans. Hours-long recordings are split, transcribed in parallel, and stitched back in order.
Which languages are supported?
English, Dutch, German, French, and Spanish, with the same speaker separation and timestamps in each.
What does it cost?
It starts free: 3 uploads a month, no credit card. Paid plans start at $10 a month and raise the limits; every plan includes the transcript plus the summaries and action items built from it.
Is my audio safe?
Your recordings are never used to train AI models and never sold. Audio stays in your account until you delete it, and deleting a conversation removes it everywhere.
Stop retyping what was already said.
Upload one recording. If the transcript is not the cleanest you have had, delete it and walk away. You have lost nothing but the upload.
3 free uploads a monthNo credit cardTranscript in minutes
