Teaspoon AI has joined the NVIDIA Inception Program

Teaspoon AI has joined the NVIDIA Inception program. It is NVIDIA’s program for startups, open to companies working in AI, data science, and accelerated computing. Members get the latest SDKs and model libraries, preferred pricing on hardware and software, free cloud credits, training through NVIDIA’s courses, and co-marketing support. It is the standard path for startups that build on NVIDIA’s stack.

Teaport is our voice appliance. It runs speech recognition and speech synthesis entirely on a Jetson Orin Nano, a board that costs a few hundred dollars. Everything you say is transcribed, answered, and spoken back on the board itself, so audio never leaves the room unless you send it somewhere. The language model behind it is any OpenAI-compatible endpoint you point it at, on the same network or in the cloud. Teaport runs in real time, handles barge-in, and treats privacy as a property of the hardware rather than a clause in a policy.

For Teaport, the program means the latest SDKs and model libraries for the chip it runs on, and preferred pricing on the hardware itself. The bet behind Teaport is that good speech can run on hardware our customers already own. This program makes that bet easier to keep.