Deepgram, the San Francisco voice AI infrastructure company, is acquiring Daily, the real-time audio and video infrastructure company that created and maintains Pipecat, an open-source framework for building voice agents. The companies announced the deal on October 29, 2026, at Deepgram's Speak event.
Neither side disclosed the price, TechCrunch reported, and the announcements did not give a closing date. Daily's team will join Deepgram, and co-founder Kwindla Hultman Kramer becomes Deepgram's head of open source.
Pipecat will stay free and open source under its existing BSD V2 licence, and contributors and users are setting up an independent Pipecat Foundation to oversee it. Deepgram said it will sell, support and keep developing all of Daily's current products and services. "For Daily's customers and the Pipecat ecosystem, nothing will change," Kramer wrote in a post on Daily's blog.
What Daily built
Kramer founded Daily with his co-founder Nina to build real-time audio and video infrastructure for calls between people. The company sells developer APIs for live audio and video, running on what Deepgram's announcement describes as a global network of 75 points of presence designed for low latency.
Daily started experimenting with voice agents in 2023. Kramer told TechCrunch the work began around the time GPT-4 launched, when the company started testing how AI could talk to people over voice and text. That led to Pipecat, which Daily open-sourced in 2024.
Pipecat handles the moving parts of a voice agent: audio and video streams, speech-to-text, a large language model and text-to-speech, while keeping delay low. According to Daily's post, the framework now has more than 200 integrations and almost 400 contributors, and was built by a core group of 18 engineers. Developers can combine models, tools, transport and APIs from different providers.
"Because of where we came from, we just always assumed real-time," Kramer told TechCrunch.
“Keeping Pipecat open was fundamental to our decision to join Deepgram.”
Why Deepgram is buying
Deepgram sells speech-to-text, text-to-speech and speech-to-speech models through cloud APIs and in self-hosted and on-premises deployments. It says it serves more than 5,000 organisations, and that more than 650,000 developers have used it to process more than 200,000 years of audio. It reached a valuation of more than $1 billion earlier this year with a $130 million Series C round, TechCrunch reported.
The two companies have worked together for years. Deepgram has supplied the transcription inside Daily's audio and video APIs since 2019, according to Daily's post, and both went through Y Combinator in 2016.
Pipecat will become the base of Deepgram's coming Voice Agent Platform, which chief product officer Ed Anuff previewed at the Speak event. The platform pairs Deepgram's models with Pipecat and Daily's infrastructure, while still letting customers use models from other providers.
Deepgram chief executive Scott Stephenson told TechCrunch the company had long wanted to work in open source and saw agent orchestration as a natural place to do it. "We want to open up the ecosystem so everybody can build voice agents," he said.
Keeping Pipecat neutral
Both companies stressed that Pipecat will stay vendor-neutral. Kramer wrote that the new foundation will be run by "a steering committee that represents the entire realtime AI ecosystem", including companies that compete with Deepgram.
"Scott was super committed to Pipecat being open source and vendor-neutral forever," Kramer told TechCrunch. Stephenson said Deepgram wants developers to keep using partners' models while it works on improving its own in those areas.
Alongside the deal, Deepgram launched the Real-Time Context Standard, an open protocol it developed with the Pipecat team. It defines metadata that voice agents can pass between models: context taken from the audio, such as a whisper or a word that was hard to make out; instructions for how a language model should use those signals; and guidance on how the reply should sound. Adapters translate these signals for each model.
The founders and the funding
Daily has raised more than $62 million, according to PitchBook data cited by TechCrunch, from investors including Tiger Global, Highfield Capital, Nat Friedman, Lachy Groom and Kevin Weil. Its largest disclosed round was a $40 million Series B in November 2021 led by Renegade Partners, which the company said at the time brought its total funding to more than $60 million.
In his post, Kramer thanked the leads of Daily's rounds, TenOneTen, Root Ventures, Freestyle, Lachy Groom and Renegade Partners, and board members Lee Edwards and Roseanne Wincek.
The deal comes as money pours into voice AI. TechCrunch noted that model makers ElevenLabs, Gradium and Fish Audio have raised large sums, and that infrastructure companies LiveKit and Vapi raised multimillion-dollar rounds this year.
