Flownote

Keep up when the language changes.

Follow multilingual meetings as they happen with automatic language identification, speaker-separated transcription, and live translated text beneath the original words. On desktop, no separate bot joins the call.

  • Automatic language identification
  • 60+ transcription and translation languages
  • No bot joins desktop calls

Short answer

Multilingual transcription turns speech in more than one language into written text. Flownote can identify spoken languages during a desktop recording, separate speakers, and show a one-way live translation in a language you choose, so a multilingual team can follow the same conversation more easily.

How it works

From conversation to a record you can use

  1. 01

    Start the meeting

    Record an online call or an in-person conversation from Flownote. On desktop, the app captures audio from your device instead of adding a meeting bot.

  2. 02

    Choose one translation language

    Select the language you want to read. Flownote keeps the original transcript visible and adds a translated line beneath each speaker as the conversation moves.

  3. 03

    Review the original record

    After the meeting, review the original-language transcript and the structured notes, decisions, and follow-ups. Confirm important names, numbers, and commitments before sharing them.

What you get

More useful than a folder of raw recordings

Flownote gives each conversation a structure you can review, correct, and return to.

  • Automatic language identification

    Flownote identifies spoken languages during the stream, including conversations where the language changes.

  • Speaker-separated transcript

    Keep each contribution connected to its speaker instead of reading one undifferentiated block of text.

  • Original and translation together

    Read the original words and a one-way translated text line beneath them while the meeting is happening.

  • Online and in person

    Use the same workflow for desktop calls and multilingual conversations happening around a table.

  • No desktop meeting bot

    Flownote captures computer audio directly, so another participant does not appear in the call.

  • Useful notes afterward

    Return to the original transcript, structured notes, decisions, and follow-ups after the conversation ends.

Original + translation

See each speaker's words—and the translation beneath them

This fictional example illustrates the current live desktop experience. It contains no customer recording or private meeting data.

Fictional product example

Launch planning · translating to English

Languages detected automaticallyEnglish translation
  1. Elena

    Spanish detected

    La prueba con clientes empieza el jueves.

    Translation

    The customer test starts on Thursday.

  2. Marc

    French detected

    Je partagerai la nouvelle maquette avant midi.

    Translation

    I’ll share the new mockup before noon.

  3. Elena

    Spanish detected

    Perfecto, entonces revisamos los resultados el viernes.

    Translation

    Perfect—then we’ll review the results on Friday.

01

Language changes

Let the conversation switch languages without losing the thread

A multilingual meeting does not always stay in one language. Flownote's current desktop transcription uses automatic spoken-language identification, so the transcript can follow language changes inside the conversation rather than requiring every speaker to stay in one preset language.

The current desktop implementation uses Soniox stt-rt-v5 with language identification and speaker diarization enabled. The app offers 60 translation targets, while Soniox documents real-time transcription and translation across 60+ languages.

02

Across meeting apps

Use one desktop layer instead of a different translator for every call

Google Meet, Microsoft Teams, and Zoom each offer their own caption or translation options, with different languages, controls, plans, and add-ons. Flownote works from the desktop audio layer, so the same transcription workflow can sit alongside different meeting apps without a separate bot participant.

That does not make Flownote a native integration with those platforms, and it does not replace an interpreter. It gives participants translated text they can follow during an ordinary meeting workflow.

03

After the meeting

Live translation helps now. The original transcript remains the record.

Live translated lines are designed to help you follow the conversation while it is happening. In the current desktop workflow, the finalized transcript and post-meeting notes are based on the original-language transcription; do not assume the translated lines will be stored or exported afterward.

Automated transcription and translation can still get names, numbers, specialized terminology, and implied meaning wrong. Review consequential details and use a qualified human interpreter for legal, medical, safety-critical, or other high-stakes conversations.

04

Responsible capture

No bot does not mean no disclosure

Flownote's desktop capture keeps another bot participant out of the call, but everyone should still know when a conversation is being recorded, transcribed, or translated.

Follow the policies and laws that apply to the participants and location, obtain any required consent, and provide a non-recorded alternative when the situation calls for one.

Questions

Frequently asked questions

What is multilingual transcription?

Multilingual transcription converts speech in multiple languages into written text. A multilingual system can identify or handle more than one spoken language instead of requiring an entire conversation to remain in a single preset language.

Can AI transcribe a meeting where people switch languages?

Yes. Flownote's current desktop transcription uses automatic spoken-language identification and can follow language changes within a stream. Accuracy still varies with audio quality, overlapping speech, accents, names, and specialized vocabulary.

Can Flownote translate a Zoom, Google Meet, or Microsoft Teams meeting in real time?

On desktop, Flownote captures computer audio and can show one-way translated text alongside the original transcript while a call is happening. It works at the device-audio layer rather than as a native Zoom, Google Meet, or Teams integration.

Does Flownote identify speakers and languages automatically?

Yes. Speaker diarization and spoken-language identification are enabled in the current desktop real-time transcription workflow.

Can I see the original transcript and translation together?

Yes. During a desktop recording, Flownote can keep the original text visible and add a translated text line beneath each speaker in the target language you selected.

How many languages does Flownote support?

The current desktop translation picker offers 60 target languages, and the underlying Soniox real-time transcription and translation documentation describes support across 60+ languages. Language availability can change as the implementation evolves.

Is the translated transcript saved after the meeting?

Do not assume so. Live translation currently helps you read the conversation as it happens, while the finalized transcript and post-meeting notes are based on the original-language transcription.

Is live translation the same as a human interpreter?

No. Flownote provides automated translated text, not human interpretation or speech-to-speech audio. Use a qualified interpreter for high-stakes conversations where nuance and verified accuracy are essential.

Follow the whole conversation—even when the language changes

Download Flownote for desktop and try live multilingual transcription with a representative, consented meeting.