New Tools Help Apps Listen and Speak

Vercel and ElevenLabs are bringing more sound tools to AI Gateway. The service lets developers use different computer models from one place. A model is a computer tool that works with information. There are 14 ElevenLabs models for speech, music, and other sounds. They cost 50% less for the first 14 days, through November 9th. Developers can also check use and costs together in AI Gateway.

Some tools turn a person's speech into written words for an app. Other tools read an app's written answer out loud for users. The voice can start before the whole answer is ready. This means users do not need to wait for every word. Developers can change one part without changing all the other parts. They can choose separate tools for listening, answering, and speaking.

Another choice puts the whole spoken conversation in one model. A user speaks, and the model answers back during the same conversation. This can help with customer support or an assistant inside an app. For example, a customer might ask to change a delivery address. The assistant can then ask for the order number to continue.

The service also lets developers make music from a written description. They can choose the style, length, and whether the music includes singing. Other tools make short sounds, such as footsteps or falling rain. Developers can give different voices to people in one recorded conversation. These choices bring several kinds of sound into the same service. Teams can use them while following their other model use and costs.