Skip to main content
BlogNewsletterRSS
Back to changelog

Use GLM-5.3-FlashX and GLM-5.3 through AI Gateway

Avatar for Nijiko Yonskai

Nijiko Yonskai

AI Gateway now supports GLM-5.3-FlashX through Z.ai and TokenHub, plus GLM-5.3 through Mistral. Try them in the playground, then connect your application through the SDK you already use.

With ngrok-managed inference, you don’t need a separate account or API key for each provider. Use your ngrok access key and credits to send requests, with authentication and billing handled by ngrok.

GLM-5.3-FlashX is the high-speed serving variant of GLM-5.3-Flash. It’s a separate model, so select the FlashX entry to try it. Check the provider’s model details for pricing and supported request formats.

Get started

  1. Open the Playground and try GLM-5.3-FlashX from Z.ai or TokenHub, or GLM-5.3 from Mistral.
  2. Check that your access key allows the provider and model you want to use, and confirm your account has credits.
  3. Follow the setup guide to choose your SDK and language, then use the integration instructions to connect your application.