Croatian transcription that holds up.
Croatian gets treated as a rounding error by most transcription tools. Canto measures it, publishes the result, and lets you check the claim on your own audio.
≤5%
Canto AI
14.5%
Industry-standard engine
2.9× fewer errors · word error rate on FLEURS, lower is better
How Canto handles this language
Because Croatian sits so close to Bosnian and Serbian, engines trained mainly on the wider region tend to blur the three together, producing text that is understandable but wrong in exactly the places a reader notices. Canto targets Croatian directly and holds the standard forms and vocabulary rather than drifting toward a neighbouring variant partway through a meeting.
No public per-language figure yet for this language; the current model sits in the top accuracy tier (under 5% error), which is what we cite here.
What the benchmark measures
Word error rate counts every insertion, deletion and substitution against a human reference transcript, then divides by the number of words spoken. Lower is better, and 5% means roughly one wrong word in twenty. We measure on FLEURS, a public benchmark of read speech across many languages, so anyone can repeat the test rather than take our word for it.
How we measure
The same audio goes through both engines with no per-language tuning, and the output is scored with the same normalisation on both sides. We publish the competing engine's figure alongside ours rather than only our own, because a number with nothing to compare it to says very little.
What you get
- Automatic language detection, including conversations that switch language mid-sentence
- Speakers separated automatically, so the transcript reads like a script
- AI titles, summaries and action items written in the language you choose
- Search and chat across everything you record, with answers that cite the moment
- A notetaker that joins Teams, Meet and Zoom, plus phone and watch recording in person
Questions
How accurate is Croatian transcription?
Canto's current model sits in the top accuracy tier for Croatian, under 5% word error rate. The engine most other notetakers build on scores 14.5% on the same public benchmark.
Does it handle mixed Croatian and English?
Yes. Detection is automatic and works mid-sentence, which matters in meetings where the technical vocabulary stays in English.
Can I get summaries and notes in Croatian?
Yes. The interface and all AI-written output (titles, summaries, action items, chat) is available in Croatian.
Try it on your own audio.
300 minutes a month, free forever. The fastest way to check an accuracy claim is to run your own recording through it.