Skip to main content
The SLNG Model Gateway is the fastest single entry to every STT and TTS model we offer. We have built this gateway specially to cover all Voice use cases. If you want to use a model for agents or any other use case you connect once, and the gateway gets each request to the right model, in the right place, as fast as possible.

​
What it does

  • One way in to many models. STT and TTS from many providers are all available through one low latency focused gateway.
  • SLNG-hosted and third-party models side by side. The gateway sits in front of both SLNG-hosted models and external providers, so you get one experience whichever model they pick.
  • Compliance and location. You call the endpoint in the region that you prefer. This request is processed in that same region.
Everything about the models can be found here.

​
Why it’s fast

  • It runs close to the user. The gateway is deployed across 14 SLNG regions. Less distance means less latency and more natural conversations.
  • It’s built for live voice. Voice conversations are continuous connections, not quick one-off requests. The gateway is built specifically for that pattern, so performance stays steady through long calls and at high volume instead of slowing down over time.
  • It adds almost nothing on top. The gateway is a thin, efficient layer, so customers get the benefits of a unified entry point without paying for it in response time.

​
Why it matters

In voice, every millisecond counts, and delays add up across a conversation. A gateway that is close to the user, chooses the best path, and holds up under long, heavy traffic keeps response times low and consistent, wherever the customer is.

​
Regional Endpoints