Definition
Call-translation latency is the time between the moment a speaker finishes a sentence and the moment the other party hears the translated version, ideally under 1.5 seconds.
In a translated call, latency is the end-to-end delay from the moment a speaker finishes a sentence to the moment the other party hears it in their language. It is the sum of five smaller delays. Audio capture and encoding takes about 30 ms, and network transit to the server 50 to 120 ms depending on region. Speech-to-text finalisation takes 200 to 400 ms, machine translation 100 to 200 ms, and text-to-speech first-audio another 200 to 400 ms. ChromePing's observed average for European language pairs is 420 ms in 2026; anything over 1.5 seconds total starts to feel non-conversational, so each stage has its own budget enforced by the pipeline.
Related terms
Speech-to-text (STT)
Speech-to-text turns spoken audio into text in near real time, the first stage of translated calls.
Text-to-speech (TTS)
Text-to-speech is the last stage of a translated call, it reads the translated text aloud in the target language's voice.
Audio codec
An audio codec compresses voice into a bitstream for transmission. Opus is the browser-standard codec used by WebRTC and ChromePing.