Realtime voice, speech, and transcription now supported on AI Gateway

Short summary

Vercel AI Gateway now supports real-time voice agents, text-to-speech, and speech-to-text transcription with the same spend controls and bring-your-own-key support as other models. Voice conversations run in near-zero latency with tools callable mid-conversation. Developers can build using a server token route and browser component, or test instantly in the interactive playground.

• •Realtime voice agents listen, respond, and speak back with low latency
• •Same cost controls, observability, and BYOK support as text/image models
• •Implement via server token route + browser component or test in playground immediately

Generated with AI, which can make mistakes.