Live Translate

1 min read

Live Translate is the one tool for live translation. Leave Source on Any language (Auto-detect) to translate one or many spoken languages into one Target, or choose the other person's language to work between exactly two languages. Start talking, and change languages while the session keeps running.

One tool, two situations

Live Translate adapts to the conversation. Leave Source on Any language (Auto-detect) when people speak one or many languages: Saydi translates all of them into one Target. Choose the other person's language when exactly two languages go back and forth. You can change it anytime.

Two simple setups

Many languages → one? Leave Source on Auto-detectGood for a presentation, lecture, or a room where people switch languages. Saydi detects each spoken language and translates all of them into the one Target you choose.See the guide Exactly two languages? Choose the other languageGood for 1-1 meetings, negotiations, or two groups. Choose the other person's language; Saydi translates both directions in the same two panels.See the guide

Which setup fits?

Setup	Any language → one	Two languages, back and forth
Languages	One or many spoken languages, all translated into 1 Target.	Exactly 2 languages: Source and Target.
Source & Target	Source: Any language (Auto-detect). Target: the one language people read.	Source: the other person's language. Target: your language.
On screen	Two panels from the start: original on the left, translation on the right. Every spoken language uses the translation panel.	The same two panels: each language can appear on either side as people take turns.

When people speak one or many languages

Leave Source on Auto-detect. Every spoken language is translated into the Target you choose.

Open Live Translate. Leave Source on Any language (Auto-detect).
Choose the Target language people will read.
Pick the microphone and start talking.
If someone switches language, Saydi keeps translating the new language into the same Target. The transcript stays.

On a big screen

Use text size 6 or more on a projector and check it from the back row. If the audience is large and needs several languages, use Event Mode so each person reads on their own phone.

Auto-detect: any spoken language becomes one target

When exactly two languages are in use

Choose the other language as Source. Both directions are translated in the same two panels.

Open Live Translate and change Source from Any language (Auto-detect) to the other person's language, for example English.
Choose Target, the language you want to read, for example Vietnamese.
Pick the microphone and start talking.
Saydi detects which of the two languages is spoken; the original and translation fill the two panels.
If a third language appears, switch Source back to Any language (Auto-detect) for one-way translation.

Voice in the same room

Turn Voice off when everyone can hear each other, so the spoken translation does not talk over the person speaking.

Two languages: change Source before Start; both directions translate

Change setup while recording

The Source and Target pickers stay available during the session. Changes apply without restarting the recorder or transcript.

Open the Source or Target picker in the top bar.
Choose a language. A specific language translates both directions; Auto-detect translates into one target.
Keep talking. History records the final setup when you click Done.

When Saydi suggests a different direction

Two examples. You are translating many languages into one: if Saydi hears exactly two languages being used, it may ask "Translate both ways?" You are translating both ways: if it hears only one language for the first 5 minutes, it may ask "Translate into one target?" The banner closes after 5 seconds and never stops the recording.

Switch

Applies the suggested translation direction without stopping the recording.

Not now

Keeps the current direction. The suggestion will not appear again in this session.

These are only suggestions. The recording continues either way.

Controls while running

Control	Use when
Sound bars	Check that they move when someone speaks. If they do not, Saydi is not hearing the room.
Pause	A break, or a private conversation you do not want in the transcript.
Done	The meeting has ended. The transcript and summaries are saved.
Reset	You intentionally want to clear the screen and start the text again.

When you finish

Click Done before closing the browser tab. The transcript and summaries are saved in History. After the session, open History and transcripts to create a Quick Summary, or find, copy, share, or export the session. Use AI Chat & Quick Summary to ask questions while the meeting is still running.

During and after the session

AI Chat & Quick SummaryAsk about the meeting while it is happening. Quick Summary is created from the saved session in History.History & transcriptsFind, copy, share, or export a past session.Plans & billingPAYG top-ups, credit balance, and invoices.

Common questions

No. Leave Source on Any language (Auto-detect) to translate any spoken language into one Target, or choose a specific language when exactly two languages go back and forth. You can change it later.

No. The recorder, connection and transcript continue. The new setup applies from that point.

Yes. Set Source back to Any language (Auto-detect) to translate any language into one Target, or choose another language for exactly two languages.

Dismiss the banner. You can change the languages yourself.