Deepgram acquires daily to expand voice AI infrastructure

Deepgram has acquired Daily, the creator of the open-source voice AI framework Pipecat, to enhance its voice AI infrastructure. The deal, which was announced on Thursday, does not disclose the financial terms but confirms that Pipecat will remain open-source. Pipecat manages voice and video streams, integrating speech-to-text, large language models, and text-to-speech technologies with reduced latency. Daily, which has raised over $62 million in funding, was co-founded by Kwindla Kramer. Kramer highlighted Deepgram’s commitment to open-source and the alignment of engineering cultures as key factors in the acquisition. Deepgram’s CEO, Scott Stephenson, emphasized the potential of Pipecat to enable developers to build expressive voice agents and end-to-end speech systems. The acquisition also allows Deepgram to explore real-time communication beyond voice through Pipecat. The companies are also launching the Real-Time Context Standard (RCS), an open protocol for voice agents co-developed with the Pipecat team. The voice AI sector has seen significant investment, with companies like ElevenLabs, Gradium, and Fish Audio raising millions, while infrastructure firms such as LiveKit and Vapi have secured multi-million-dollar funding rounds this year. Kramer noted that Pipecat was initially envisioned as a toolkit for AI-native software rather than just a voice toolkit. Stephenson added that the acquisition supports Deepgram’s goal of fostering an open ecosystem where developers can use the best models for their use cases.