Developer Builds Voice Agent From Scratch Using Claude Code And AssemblyAI API
#1802Santiago@SVPINOThe Voice Agent API is a complete STT → LLM → TTS pipeline.
Everything happens through a single connection to a single API. You don't need a separate LLM, an STT API, a TTS API, etc.
The API handles interruptions, lets you call tools, and gives you access to a bunch of different voices.
I built something like this about a year ago, and it took quite a bit of time and 4 different API Keys to different services.
You can now do it all with a single pipeline.
Here is the link to the playground if you want to give this a try: https://www.assemblyai.com/dashboard/playground/voice-agent
Thanks to the @AssemblyAI team for partnering with me on this post.
Santiago@svpinoStep-by-step video to build a voice agent from scratch. I'm doing this using Claude Code, because writing code by hand in 2026 is yucky. I'm using @AssemblyAI's new Voice Agent API, which is all you need to get this done, instead of stitching together 12 different components.
#1802Santiago@SVPINOHere is the YouTube link to the video: https://youtu.be/z-sqQeVmSf8
Santiago@svpinoThe Voice Agent API is a complete STT → LLM → TTS pipeline. Everything happens through a single connection to a single API. You don't need a separate LLM, an STT API, a TTS API, etc. The API handles interruptions, lets you call tools, and gives you access to a bunch of different voices. I built something like this about a year ago, and it took quite a bit of time and 4 different API Keys to different services. You can now do it all with a single pipeline. Here is the link to the playground if you want to give this a try: https://www.assemblyai.com/dashboard/playground/voice-agent Thanks to the @AssemblyAI team for partnering with me on this post.