Scribix
AI transcription that turns video and audio recordings into speaker-labeled, word-timestamped text, with a one-time 45-minute free allowance to try it.
🔗 Visit Scribix
Description
Turning a recorded conversation into usable written text has traditionally meant either paying a human transcriptionist by the minute or fighting with free tools that garble speaker changes and lose timing. Scribix targets that specific pain point directly: accurate, speaker-labeled transcripts with timestamps, so it's clear who said what and exactly when.
It transcribes video (MP4, MOV, WebM, AVI) and audio (MP3, WAV, M4A) files, with automatic language detection and speaker identification, plus word-level timestamp synchronization for precise playback navigation. Output can be exported in five formats: TXT, DOCX, SRT, VTT, and CSV. The Free plan gives 45 minutes of transcription as a one-time lifetime allowance (not a monthly renewal), using the standard transcription model, with uploads up to 10 hours long. The Pro plan is $10/month ($120/year) for 2,400 minutes monthly (40 hours), the best-available transcription model, AI-generated notes, translation, unlimited exports, and uploads up to 5GB video / 1GB audio.
💬 Our review
The short version: at $10/month for 40 hours of transcription, Scribix's Pro tier is aggressively cheap for anyone doing bulk interview or podcast transcription — but the "free" plan is really a one-time 45-minute trial, not an ongoing free tier, so don't count on it for continuous light use.
Against Otter.ai or Rev, which mix AI and human transcription at higher per-minute costs, Scribix undercuts on price if the automated speaker labeling is accurate enough for your use case — that's the one thing to verify with your own recording before committing, since speaker-ID accuracy is where automated transcription services vary most in practice. Descript goes further by bundling actual video editing on top of transcription, which Scribix doesn't attempt. Best fit: content creators, journalists, and researchers doing regular interview transcription who want low-cost bulk minutes. Weaker fit: anyone needing guaranteed near-perfect accuracy (legal, medical) — human-reviewed services remain safer there.
💰 Pricing
📊 Global score
🤖 AI-enriched data
Free : 45 minutes à vie (allocation unique, pas mensuelle), modèle standard, upload jusqu'à 10h. Pro : 10 $/mois (120 $/an), 2 400 minutes/mois (40h), meilleur modèle disponible, notes IA + traduction, exports illimités, upload jusqu'à 5 Go vidéo / 1 Go audio.
Pros
Prix très bas pour un gros volume (10 $/mois pour 40h)
Labellisation automatique des locuteurs avec horodatage précis
5 formats d'export (TXT, DOCX, SRT, VTT, CSV)
Détection automatique de la langue
Cons
Palier gratuit = essai unique de 45 min, pas un vrai palier récurrent
Précision de la labellisation locuteurs à vérifier sur ses propres enregistrements
Pas d'édition vidéo intégrée (contrairement à Descript)