Create an AI Assistant

Create a new AI assistant for conversations, support tickets, AI chat, or voice calls.

Endpoint

POST

/ai-assistants

json

Parameter	Type	Required	Description
`organization_id`	Number	Yes	ID of the organization that owns the assistant
`name`	String	Yes	Assistant display name
`assistant_type`	String	Yes	Accepted values used by the configuration UI: `conversation`, `support-ticket`, `ai-chat`, `voice`
`ai_model_id`	Number	Yes	AI model ID. Use the AI Models API to list valid models
`connection`	String	No	Model credential source: `system`, `organization`, or `specific`
`specific_openai_secret_key`	String	No	Provider key used only when `connection` is `specific`
`timezone`	String	No	IANA timezone used by scheduling and assistant context
`instructions`	String	No	Advanced-mode prompt instructions
`instructions_language`	String	No	`ENG` or `ESP`
`is_easy_mode`	Boolean	No	Whether the assistant uses guided mode
`easy_mode_config`	Object or null	No	Guided-mode source config. Use `null` for advanced mode
`telnyx_ai_assistant_config`	Object or null	No	Voice runtime configuration for voice assistants

Voice assistants use the same endpoint with assistant_type set to voice.

The configuration UI duplicates an assistant by creating a new assistant with the same configuration. To duplicate through the API:

Retrieve the source assistant with GET /ai-assistants/:id.
Remove server-managed fields such as id, created_by_user_id, created_at, updated_at, and deleted_at.
Change name and call POST /ai-assistants.

Please log in to report issues with our documentation.