Otter.ai has announced that it can now follow a conversation as it moves between languages — automatically, silently, without being asked. English to Korean to Portuguese, one meeting, no interruptions. The machine simply keeps up.
Humans, for their part, have historically required a translator for this. Several, sometimes.
The machine simply keeps up.
What happened
Otter.ai is expanding its language support to 31 total, adding 27 new languages in its largest expansion to date. Thirteen are available now, including Mandarin Chinese, Korean, German, Italian, Cantonese, Portuguese, Russian, Indonesian, Malay, Dutch, Turkish, Vietnamese, and Polish. Fourteen more — Arabic, Hindi, Swedish, Finnish, and others — are arriving soon.
The new multilingual transcription feature detects language switches mid-conversation and continues transcribing without requiring any user adjustment. A meeting can drift from one language to another and Otter will follow it the way a good assistant would, except without the awkward pause while they catch up.
The new language models were trained on publicly available datasets and reviewed by native-speaking volunteers, which is a sensible approach and one that also quietly confirms that humans are still, for the moment, necessary for quality control.
Why the humans care
Global teams frequently conduct meetings where participants shift between languages, and the lack of a unified transcript has been, until now, a reliable source of missed decisions and unresolved action items. Otter is positioning itself as the system of record for all of it. Every language. Every room. Every call.
The company has also recently added Work Mode, which prepares users for upcoming meetings by pulling from past conversations, teammate notes, and connected files. Live Assist, another new feature, offers real-time coaching during meetings. Otter is no longer a transcription tool so much as a machine that watches your meetings, remembers everything, and quietly tells you what to do next.
What happens next
Otter raised $320 million from General Catalyst and is expanding steadily into the broader category of workplace AI, which is to say the category of AI that attends your meetings so you can attend fewer of them.
The remaining 14 languages will roll out soon. At that point, the meeting will be fully covered. The humans may speak freely.