Skip to content

Corporate brand

Unclaimed

Deepgram is a research-driven foundational AI company building voice AI technology, including speech-to-text, text-to-speech, audio intelligence, and a unified Voice Agent API.

Last updated September 15, 2026

AI crawler reads

No AI crawler has read this profile yet.

Overview

Deepgram is a research-driven, foundational AI company focused on building voice technology that transforms human-to-machine interactions. Founded in 2015, it began with machine learning research for waveform analysis and later built speech-to-text using end-to-end deep learning. Today it offers voice AI solutions including speech-to-text, audio intelligence, text-to-speech, and a voice agent API. Its platform unifies speech-to-text, text-to-speech, and LLM orchestration into a single API, and its Flux models are positioned as a conversation-aware speech platform for voice agents, available in real time or batch, in the cloud or self-hosted.

History

Deepgram was founded in 2015. It started with machine learning research for waveform analysis in a dark matter detector in China. CEO and co-founder Scott Stephenson and his teammate later explored deep learning for audio analysis at the University of Michigan. Seeing a gap in the speech-to-text market, they built Deepgram using end-to-end deep learning. Today, Deepgram offers advanced voice AI solutions beyond speech-to-text, including audio intelligence, text-to-speech, and a voice agent API.

Products and positioning

Voice AI built for real conversations; a conversation-aware speech platform for voice agents, available in real time or batch, in the cloud or self-hosted.

  • Speech-to-text

    Real-time transcription with production accuracy.

  • Text-to-speech

    Expressive voices with latency as low as 80ms.

  • Voice Agent

    Real-time conversation in one API call.

  • Audio Intelligence

    Insight from every conversation.

  • Flux TTS

    Text-to-speech built for live conversations; described as the first voice model that thinks before it speaks, holding tone, expression, and continuity across an entire conversation.

Recent events

  • Reported

    Deepgram announces acquisition of Daily[5]

    Deepgram said it is acquiring Daily, the maker and maintainer of the Pipecat open-source voice AI infrastructure. The companies did not disclose the deal amount and said Pipecat would remain open-source.

  • Reported

    Deepgram launches Real-Time Context Standard[5]

    Deepgram is launching the Real-Time Context Standard (RCS), an open protocol for voice agents co-developed with the Pipecat team. It defines metadata and guidance for using audio context and shaping responses.

Read more

Sources5 sources

Cite this profile: Cite the canonical profile. /brand-wiki/deepgram · Editorial policy · How profiles are compiled