Live translation
Translate meetings and events as they happen, in the languages each person reads.
Live translation turns speech into text, and optionally speech, while a meeting or event is happening. You set a Spoken Language for what is being said and a Translation language for what you want to read, then start recording.
Allow live session audio#
Grant microphone access so JotMe can hear you, and screen or system-audio access so it can hear the other side of an online meeting. Without system audio, it hears only your microphone.
Heads up
On macOS, system-audio permission takes effect only after you quit and reopen JotMe.
Prepare before the meeting or event#
Starting from a connected calendar event lets you prepare notes, reserve its join link, and save its languages and audio settings before anyone arrives. The event restores that setup the next time you open it. See Calendar for preparation and Share live translation for delivery.
Note
A guest can join the shared browser link from a phone or computer. The Chrome extension is a separate Google Meet workflow; it does not join a desktop-hosted session.
Set the spoken language#
The spoken language is what people are actually saying, and it decides which model transcribes them. Click either side of the top-bar language control, choose the model, then select the spoken and translation languages in their separate columns.
Language settings
Select AI model
Spoken language
Language they are speaking
Translation output
Language you’re seeing
| Model | Best for | What it can do | Usage and tradeoff |
|---|---|---|---|
| Basic | Everyday captions, or a conversation that switches among the 10 Multilingual languages | Transcribes one selected spoken language, or automatically follows those 10 languages with Basic → Multilingual. Feeds live text translation. | 1× minutes. Standard caption and translation quality. |
| Premium Quality | English plus one selected supported language | JotMe's highest-quality captions and text translation. It always transcribes English in addition to the language selected in the picker: select Japanese, for example, and it transcribes both Japanese and English. | 2× minutes. Select the supported non-English language expected in the session. |
| Fast (Audio) | A language the other models do not cover, the lowest latency, or speech-to-speech translation | Recognises 100+ spoken languages without selecting one first. Produces translated captions and, after audio setup, one-way or two-way translated speech. | 5× minutes. Caption translation quality can be lower than Basic or Premium Quality. |
Those 10 languages are English, Spanish, French, German, Hindi, Russian, Portuguese, Japanese, Italian, and Dutch at the same time. It follows speakers as they switch languages; you do not select all 10 individually.
Pick this before you start: JotMe blocks changes to the spoken language during a recording, because the transcription stream is already running on that model.
Set the translation#
Translation has its own control next to the spoken language, with three modes:
- Single: one target language.
- Multilingual: up to 3 target languages shown at once, on Premium. See Languages for the models and what each multiplies.
- Fast (Audio): the audio-to-audio mode, available when the spoken language is also set to Fast.
Two-way sessions label the directions You hear and Others hear, and the two must be different languages.
Choose your environment#
Under Settings → Live session → Your environment, choose Online or In-person so JotMe labels speakers correctly. Configure the microphone and other session-wide preferences in Desktop app settings.
Tune translation quality#
Settings → Live session → Translation quality balances speed and context. The same section contains automatic adjustment, real-time memory, and transcription quality. See Desktop app settings for the controls and model-specific availability.
During the session#
The session window has three tabs:
- Transcript: transcription and translation text as it arrives.
- Ask: questions answered from the live session.
- Note: your own Markdown notes, enhanced against the transcript after the session.
Reverse translation can be toggled from the transcript toolbar, and the transcript window can be maximised or minimised without stopping the session.
When limits run out#
JotMe meters transcription and live translation separately. If translation runs out, it can continue with transcription only; if transcription runs out, the session stops and is saved. See Usage limits for plan allowances, multipliers, and limit behavior.
If JotMe cannot hear the meeting or event#
Check microphone and system-audio access first. On macOS, quit and reopen JotMe after granting permission. For device selection and recovery steps, use Troubleshooting.
What you get afterwards#
Stopping creates the enhanced notes and saves the session under Recordings. See Note-taking for what is generated. If you want a group chat created for attendees, enable the post-session chat before stopping; see Multilingual chat.
Choose a workflow#
Share live translation
Reserve the join link early and post it into a chat.
Speech to speech
Voice-to-voice translation, and the audio routing it needs.
In-person meetings
Translate a face-to-face conversation on one machine.
Zoom, Teams, Meet, and other online meetings
Run a session alongside the meeting tool you already use.
Events and conferences
Translate a room, with attendees reading on their own phones.
Webinars
Broadcast one speaker to an audience in several languages.
Related features#
Live translation feeds the session record and the workflows around it.
Calendar
Transcription
Note-taking
Recording
Ask JotMe
Multilingual chat