Deepgram to Acquire Daily, Maker of Pipecat Voice AI Framework
Deepgram is acquiring Daily, the company that makes and maintains Pipecat, the open-source framework for building voice AI agents. The deal was announced on October 29, 2026, according to TechCrunch. Neither company disclosed the purchase price.
Daily provides hosted infrastructure for voice and video calls, and it is the maintainer of Pipecat, which sits between real-time media transport and the model services that power a conversational agent. Deepgram and Daily said they will keep Pipecat open-source. Daily co-founder Kwindla Kramer said Deepgram CEO Scott Stephenson had committed to Pipecat remaining open-source and vendor-neutral permanently.
The two companies have a longer history than the transaction suggests. Both were part of the same Y Combinator batch in 2016. Daily has raised more than $62 million in funding to date, according to PitchBook data cited by TechCrunch.
Deepgram is also launching the Real-Time Context Standard (RCS), an open protocol for voice agents that it co-developed with the Pipecat team. The protocol is positioned as a common way for voice agent components to exchange real-time context, though the details of its message format and governance will matter more to implementers than the announcement itself.
The companies were already working together before the acquisition. Pipecat ships with native Deepgram services, and Deepgram lists Daily.co as a partner for live transcription in video calls. Daily Bots, which Daily describes as an open-source cloud for voice AI built on Pipecat, natively integrates Deepgram. On August 17, 2026, Deepgram published instructions for building a voice agent with Pipecat using Deepgram Flux STT and Flux TTS. Deepgram's own blog lists the Daily deal in its Announcements section, under the headline "Deepgram is acquiring Daily, makers of Pipecat, to offer the first complete, production-ready, open stack for conversational voice AI."
Deepgram had announced the completion of its latest funding round in January 2026, in a post titled "Today We Start the Next Chapter of the Voice AI Economy." The Daily transaction is a separate step, and the announcement does not indicate how it will be financed.
The broader context here is a consolidation of the voice AI stack. Speech-to-text, text-to-speech, the orchestration layer that routes audio and turn-taking between them, and the hosted media infrastructure that carries the call have typically been sourced from different vendors. Bringing the orchestration framework and the media layer under the same company as a major speech provider changes who controls the default path for a developer who starts with Pipecat.
In my view, the most consequential question for engineering teams is not the acquisition itself but whether vendor neutrality holds in practice. Pipecat's appeal has rested on its ability to swap providers behind a common pipeline. Developers who built on it with other speech vendors will want to see how the roadmap, the service integrations and the governance of RCS evolve after the close. The stated commitments are clear, and they are the terms against which those later releases can be measured.
For practitioners, the upside is easier to describe. A single open stack from transport to model integration could reduce the amount of glue code needed to get a low-latency voice agent into production, and an open protocol for real-time context could let components from different vendors interoperate. Whether that materializes depends on adoption of RCS beyond the two companies that co-developed it, and on the open-source project continuing to attract contributors who are not on Deepgram's payroll.
