Microsoft said the new ‘consecutive interpretation’ mode is designed for interactive, two-way conversations across a language pair.
Microsoft is introducing a new mode, called “consecutive interpretation,” for its Interpreter agent in Teams.
The way it works is pretty straightforward: You click the “More” menu during a call and select “Turn on Interpreter.” From there, a new option for “Consecutive interpretation” appears, where you choose the two specific languages for the conversation and turn the feature on.
Once enabled, the Interpreter agent actually joins the meeting and appears on stage for everyone to see. It creates a structured, turn-based flow where one person speaks, pauses, and then the agent delivers the translation to the entire group before the next person can respond.
Microsoft said that consecutive interpretation is available now for meeting organizers using the Teams Public Preview, provided they have a Microsoft 365 Copilot license. For now, it supports any language pair between a set of 10 languages, including Chinese (Simplified), Chinese (Traditional), English, French, German, Italian, Japanese, Korean, Portuguese, and Spanish.

Interpreter was rolled out as a public preview for Teams Meetings in early 2025. It is basically an AI agent that provides real-time speech-to-speech (STS) interpretation during meetings and voice calls.
The system has a neat little “voice simulation” capability that can replicate the original speaker’s tone, pitch, and intonation, instead of using the robotic, generic system voice these kinds of systems are known for. It also has an audio balance slider that lets a user choose to hear only the translated AI voice or blend it to hear a faint background of the original speaker’s voice.
Before this latest “consecutive interpretation” update, the agent had just one mode: real-time simultaneous interpretation, which made it suitable for broadcast-style meetings, large presentations, and scenarios where one person speaks for a prolonged period, and the other meeting participants are mostly just there listening.

