Telnyx AI

Connect Telnyx AI to use Telnyx-hosted models for AI chat, Message Templates, and flow blocks. Telnyx AI generates or analyzes text; a separate email or text-message integration sends completed messages.

  • You want to use Telnyx-hosted models for AI chat, Message Templates, or flow-based generation and analysis.
  • You want to choose from the Telnyx-hosted models available to your account.
  • Confirm that you have a Telnyx account and an active Telnyx API key that can access Telnyx inference.
  • Review the Telnyx Inference documentation if you need provider-side information about Telnyx models or access.
  • Confirm that Use AI on this account is on and that your role can create integrations. While AI is off, the integration list is hidden; without create permission, New Integration is not shown.
Choose Telnyx AI (Beta) in the AI integration wizard.
  1. Open Integrations → AI.
  2. Click New Integration and choose Telnyx AI.
  1. Review the Name, which defaults to “Telnyx AI,” select Enter new credentials, and enter your Telnyx API key.
  2. Select Verify API Key. This saves the integration, asks Telnyx to validate the API key by requesting the models available to your account, and loads the Telnyx-hosted models.
  3. After validation succeeds, optionally choose a Default Model, then select Save & Continue.
Review the blank Telnyx AI connection fields before verifying an API key.
  1. On the Verify step, review Credentials stored. Before the second check, it confirms only that credentials are saved. To run another live Telnyx validation, select Check credentials. After it succeeds, the status reports “These credentials work.” Then select Done.

Default Model appears after Verify API Key succeeds. It lists only models that Telnyx identifies as hosted on its own infrastructure. AI uses an available Llama model by default when one exists; otherwise, it uses the first available Telnyx-hosted model. Availability depends on your Telnyx account.

In the Flow Editor, use Model (optional) to override the integration’s default for one AI block.

Note
AI chat requires a model that supports tool calling. Support varies across Telnyx-hosted models. If Telnyx AI rejects a request, choose a different model from the Model menu in AI chat or review the model information in Telnyx.
  • In AI chat, choose Telnyx AI as the provider and select a tool-capable model under Model. Ask about your account or the documentation; suggested changes wait for your approval.
  • In Message Templates, open AI Assistant, choose Telnyx AI, and ask for a draft, rewrite, or tone change. Apply or dismiss each proposed change from its preview.
  • In the Flow Editor, select the Telnyx AI integration to generate message text or analyze content. A separate email or text-message integration sends completed messages.
The provider does not validate
  • Confirm that the Telnyx API key is correct, active, and able to access Telnyx inference.
  • Use the returned error to distinguish an authentication problem from an account-access or provider-availability problem.
No model choices appear
  • Confirm that validation succeeded. The same model-list request validates the credential and populates the selector.
  • Check that the Telnyx account has access to at least one inference model.
  • Select Check credentials in the wizard’s Verify step to refresh the model list immediately. Otherwise, AI can reuse the cached list for up to 24 hours.
An AI action fails after setup
  • Confirm that the selected model is still available to the Telnyx account.
  • Review the returned error and test again with a short, representative input.
  • Intermittent failures may be caused by Telnyx rate limits or temporary service problems.
  • AI: review AI provider concepts and supported workflows.