In Arabic: زمن الاستجابة (Latency)
APIs and techThe time between sending a request and receiving the first piece of audio. It barely matters for pre-made files such as ads, but voice assistants and live calls need it very low to avoid awkward silence.
A way for your software to send text to a voice service programmatically and receive audio back, without using the web interface by hand. It suits generating thousands of messages or voice inside an app or phone system.
Sending audio in chunks as it is generated rather than waiting for the whole file, so playback starts almost immediately. Essential for live conversation, voice assistants and long texts.
A secret token that identifies your account when calling an API and against which usage is billed. Keep it on your server, never in app or website code users can see.
The maximum number of requests you may send to an API in a given period. Exceed it and the server usually answers with HTTP 429, so you retry after a short wait.
A URL on your system that a service calls automatically when a long task finishes, instead of your system asking every few seconds whether it is done. Common for slow jobs such as dubbing and video generation.
A system that converses with users by voice in real time, combining speech to text, a language model that understands and replies, and text to speech. It needs low latency and a solid grasp of the caller's dialect.
Write your script, pick the dialect and voice, and download the file in minutes. Start free; paid plans from $1/month.