Deepgram’s new Aura software turns text into generative audio with a “human-like voice.” The 9-year-old voice recognition company has raised nearly $86 million to date on the strength of its Voice AI platform. Aura is an extremely low-latency text-to-speech voice AI that can be used for voice AI agents, the company says. Paired with Deepgram’s Nova-2 speech-to-text API, developers can use it to “easily (and quickly) exchange real-time information between humans and LLMs to build responsive, high-throughput AI agents and conversational AI applications,” according to Deepgram.

In an interview with TechCrunch, Deepgram co-founder and CEO Scott Stephenson emphasized that great voice models have been expensive and slow to compute, while cheaper, faster ones sound robotic.

“Deepgram’s Aura combines human-like voice models that render extremely fast (typically in well under half a second)” and does so at a low price, notes TechCrunch.

Deepgram is inviting people to test Aura for free, along with its Nova-2 speech-to-text.

Deepgram claims Aura’s pricing “currently beats virtually all its competitors at $0.015 per 1,000 characters,” but just barely. Google’s WaveNet pricing is 0.016 per 1,000 characters, the same as Amazon’s Polly neural voices. “Amazon’s highest tier, though, is significantly more expensive,” TechCrunch reports.

“Voice will become the predominant way we interact with technology — and for that to work, AI systems must be highly tuned to enable natural conversation at scale,” Deepgram explains in a press release.

No mention is made of languages other than English in which Aura can generate speech. However, the company does stipulate that it can speak based on “any text input, including responses from LLMs like ChatGPT, in fractions of a second,” for fluid, complex conversations.

Related:
Clarifai and Deepgram Announce Strategic Partnership to Accelerate AI Innovation, PR Newswire, 3/18/24