Translation glossary

Letting a speaker interrupt playing audio, so a conversation can be interrupted naturally instead of waiting.

Alternating between two or more languages within a single conversation.

Interpreting after the speaker pauses, alternating turns between speaker and interpreter.

Software that holds a natural back-and-forth conversation in speech or text, understanding and responding in real time.

The documented facts an AI agent answers from, so its replies stay accurate instead of made up.

The source and target languages of a translation, e.g. Hungarian → Chinese.

The delay between speaking and hearing the translated result.

Adapting content and product experience to a specific language and culture.

Software that translates text or speech from one language to another.

Translation by neural networks that consider whole-sentence context, not word-by-word substitution.

A message crafted to override an AI agent's instructions and make it act against its intended role.

Translating speech as it is spoken, with only a sub-second delay, so a conversation flows naturally.

Interpreting a speaker continuously, almost in parallel, instead of waiting for them to pause.

The protocols and networks used to place calls to ordinary phone numbers.

Determining who is speaking and when, to separate multiple voices on a call.

Automatic speech recognition that converts spoken audio into written text.

Synthesising natural-sounding spoken audio from written text.

Translating both directions of a conversation at once, so each side speaks and hears their own language.

Reproducing a person's vocal characteristics so synthesised speech sounds like them.

Transcribing a spoken voice note, translating it, and optionally re-voicing it in the listener's language.

A browser technology for real-time audio and video without installing an app.