Deepgram is a deep learning company with a new approach to automatic speech recognition that is significantly more accurate, faster, more efficient and cost effective than existing technology.

Deepgram helps enterprises unlock the potential of their audio data with custom trained speech recognition built to scale.

Deepgram was founded in 2015 by Scott Stephenson and Noah Shutty. The company is headquartered in San Francisco, California.

Deepgram is building building the speech company. Beyond transcription, we will unlock everything that comes with “solving speech.” This means asking, truly: “What can a human do?”

Deepgram has equipped their next-gen speech recognition with features thoughtfully designed to make it the speech tool for enterprise. The company has focused on speed to support customer processing at any scale and armed their product with features like real-time streaming, multichannel, and language support.


Deepgram's custom-trained speech recognition continuously learns at scale to improve and unlock value for your business. Deepgram MissionControl includes the features, scalability, and flexibility that businesses need. No tricks or compromises involved.


Deepgram is backed by Nvidia GPU Ventures, Wing VC, Y Combinator, SAP and others. The company raised $12M in a Series A round on Mar 18, 2020. This brings Deepgram's total funding to $13.9M to date.



  • Year founded: 2015
  • Funding Info: $13.9M over 3 Rounds (Latest Funding Type: Series A)
  • Yearly Revenue: NA
  • Employee Size: 11-50
  • Business Valuation: NA
  • City/Town: San Francisco
  • State: California
  • Country: United States
Related businesses