Deepgram Acquires Daily to Strengthen Voice AI Infrastructure

Deepgram announced on Thursday that it is acquiring Daily, the company behind the widely adopted Pipecat framework. Pipecat orchestrates speech‑to‑text, large language models, and text‑to‑speech components in a low‑latency pipeline for voice and video applications.

Daily also operates a hosted infrastructure for real‑time communications, and the deal stipulates that Pipecat will remain open‑source and vendor‑neutral. Both startups share a Y Combinator 2016 cohort; Deepgram recently reached unicorn status after a $130 million Series C, while Daily has raised over $62 million from investors such as Tiger Global and Nat Friedman.

Daily co‑founder Kwindla Kramer emphasized that Deepgram CEO Scott Stephenson’s commitment to keeping Pipecat open forever aligned with Daily’s engineering culture, giving the team freedom to work across the stack within a larger organization.

Stephenson noted that the acquisition enables developers to build expressive, end‑to‑end voice agents by plugging in best‑of‑breed models through a shared context bus. Deepgram is also launching the Real‑Time Context Standard (RCS), an open protocol co‑developed with the Pipecat team.

Why it matters for GPU and AI infrastructure

The move highlights surging demand for low‑latency, GPU‑accelerated voice pipelines. An open orchestration layer like Pipecat reduces integration overhead for AI workloads running on GPU clouds such as AiGpu, allowing faster deployment of real‑time voice agents at scale.