Aura, Explained Simply
What Aura is, what it does well, and what it costs — released 2024-09-18 by Deepgram.
- Deepgram just added a new connection method (WebSocket API) for its Aura text-to-speech system.
- It makes it easier to handle people interrupting the AI and to manage lots of chats at once.
Want a heads-up when Deepgram updates?
We email you the moment a new Deepgram version drops — plus plain-English release notes like these. No spam.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
What this is and who wants it
Aura is a tool by Deepgram that turns written text into spoken audio (text-to-speech). This update is mostly for software developers who are building voice apps or AI assistants. If you are making a chatbot that needs to talk back out loud in real time, this makes the setup smoother and faster.
What it does well
Its main strength is cutting down on awkward pauses. When you pair Aura with a "large language model" (the brain of an AI chatbot), the chatbot doesn't need to finish writing a whole sentence before it starts speaking. Instead, Aura can take "text tokens" (tiny chunks of words) the exact second the AI creates them and start turning them into audio right away. This makes the conversation feel much more natural. It is also good at handling users who cut the AI off mid-sentence, and it gives developers a simple way to manage many chats at once by giving each conversation its own private connection.
Worth knowing
The main catch is that this update is all about the behind-the-scenes wiring. It uses a "WebSocket API," which is a type of live, two-way connection between a computer program and a server. To use it, developers have to set up their apps to work with this specific kind of streaming connection.
This is our plain-language summary. Read the complete, official notes on the Deepgram official changelog ↗.
Google Calendar