Talklation's face-to-face mode uses explicit speaker turns instead of trying to guess who is talking. Place the phone where both people can see it, choose the two languages, and tap the side of the person who will speak next. The app then keeps that source-to-target direction clear for the current turn.
Manual turns reduce a common source of confusion in bilingual conversations: an automatic system can switch direction at the wrong moment when voices overlap, someone pauses, or background speech is present. With Talklation, the selected speaker is visible before listening starts, so both people know which language the app expects.
Prepare both languages before the conversation if you want the core flow to work offline. Each person should try one short sentence in advance, and the phone should remain close enough to capture clear speech. For noisy environments, shorter turns and a brief pause before switching speakers usually make the transcript easier to review.
The two directions remain part of one conversation session. When the conversation ends, Talklation saves the completed text locally so you can review what each person said. If you later request an AI summary, that optional action requires an internet connection and sends the completed record you selected rather than live microphone audio.
Face-to-face mode is useful for travel, reception desks, appointments, field work, and other situations where two people share one phone. It is not a substitute for a professional interpreter in medical, legal, safety-critical, or other high-stakes conversations where exact meaning must be verified.
If either side is not ready, stop and prepare the missing system language resource rather than continuing with an uncertain direction. Talklation's readiness checks are designed to make that limitation visible before the conversation begins.
