/telnyx-ai-assistants-python

name: telnyx-ai-assistants-python
description: >-
  AI voice assistants with custom instructions, knowledge bases, and tool
  integrations.
metadata:
  author: telnyx
  product: ai-assistants
  language: python
  generated_by: telnyx-openapi-pipeline
  contract: v2

<!-- Auto-generated from Telnyx OpenAPI specs. Do not edit. -->

Telnyx AI Assistants - Python

Installation

pip install telnyx

Setup

import os
from telnyx import Telnyx

client = Telnyx(
    api_key=os.environ.get("TELNYX_API_KEY"),  # This is the default and can be omitted
    public_key=os.environ.get("TELNYX_PUBLIC_KEY"),  # Required for client.webhooks.unwrap() (webhook verification)
)

All examples below assume `client` is already initialized as shown above.

Error Handling

All API calls can fail with network errors, rate limits (429), validation errors (422), or authentication errors (401). Always handle errors in production code:

import telnyx

try:
    assistant = client.ai.assistants.create(
        instructions="You are a helpful assistant.",
        name="my-resource",
        model="openai/gpt-4o",
    )
except telnyx.APIConnectionError:
    print("Network error — check connectivity and retry")
    raise
except telnyx.RateLimitError:
    # 429: rate limited — wait, then retry the request (bounded exponential backoff)
    import time
    time.sleep(1)  # Check Retry-After header for actual delay
    # ...retry the call here (give up after a few attempts); otherwise re-raise:
    raise
except telnyx.APIStatusError as e:
    print(f"API error {e.status_code}: {e.message}")
    if e.status_code == 422:
        print("Validation error — check required fields and formats")
    raise  # A failed request must never look like success to the caller

Common error codes: `401` invalid API key, `403` insufficient permissions, `404` resource not found, `422` validation error (check field formats), `429` rate limited (retry with exponential backoff).

Important Notes

• **Phone numbers** must be in E.164 format (e.g., `+13125550001`). Include the `+` prefix and country code. No spaces, dashes, or parentheses.
• **Pagination:** List methods return an auto-paginating iterator. Use `for item in page_result:` to iterate through all pages automatically.
• **Model availability** varies by account. If a model returns 422 "not available for inference", use `client.ai.assistants.list()` to discover working models. Commonly available: `openai/gpt-4o`, `Qwen/Qwen3-235B-A22B`.

Reference Use Rules

Do not invent Telnyx parameters, enums, response fields, or webhook fields.

• If the parameter, enum, or response field you need is not shown inline in this skill, read [references/api-details.md](references/api-details.md) before writing code.
• Before using any operation in `## Additional Operations`, read [the optional-parameters section](references/api-details.md#optional-parameters) and [the response-schemas section](references/api-details.md#response-schemas).

Core Tasks

Create an assistant

Assistant creation is the entrypoint for any AI assistant integration. Agents need the exact creation method and the top-level fields returned by the SDK.

`client.ai.assistants.create()` — `POST /ai/assistants`

| Parameter | Type | Required | Description | |-----------|------|----------|-------------| | `name` | string | Yes | | | `instructions` | string | Yes | System instructions for the assistant. | | `tags` | array[string] | No | Tags associated with the assistant. | | `model` | string | No | ID of the model to use when `external_llm` is not set. | | `tools` | array[object] | No | Deprecated for new integrations. | | ... | | | +28 optional params in [references/api-details.md](references/api-details.md) |

assistant = client.ai.assistants.create(
    instructions="You are a helpful assistant.",
    name="my-resource",
    model="openai/gpt-4o",
)
print(assistant.id)

Primary response fields:

• `assistant.id`
• `assistant.name`
• `assistant.model`
• `assistant.instructions`
• `assistant.created_at`
• `assistant.a2a_agents`

Chat with an assistant

Chat is the primary runtime path. Agents need the exact assistant method and the response content field.

`client.ai.assistants.chat()` — `POST /ai/assistants/{assistant_id}/chat`

| Parameter | Type | Required | Description | |-----------|------|----------|-------------| | `content` | string | Yes | The message content sent by the client to the assistant | | `conversation_id` | string (UUID) | Yes | A unique identifier for the conversation thread, used to mai... | | `assistant_id` | string (UUID) | Yes | Unique identifier of the assistant. | | `name` | string | No | The optional display name of the user sending the message | | `stream` | boolean | No | When true, the response is streamed as Server-Sent Events (`... |

response = client.ai.assistants.chat(
    assistant_id="550e8400-e29b-41d4-a716-446655440000",
    content="Tell me a joke about cats",
    conversation_id="42b20469-1215-4a9a-8964-c36f66b406f4",
)
print(response.content)

Primary response fields:

• `response.content`

Create an assistant test

Test creation is the main validation path for production assistant behavior before deployment.

`client.ai.assistants.tests.create()` — `POST /ai/assistants/tests`

| Parameter | Type | Required | Description | |-----------|------|----------|-------------| | `name` | string | Yes | A descriptive name for the assistant test. | | `destination` | string | Yes | The target destination for the test conversation. | | `instructions` | string | Yes | Detailed instructions that define the test scenario and what... | | `rubric` | array[object] | Yes | Evaluation criteria used to assess the assistant's performan... | | `description` | string | No | Optional detailed description of what this test evaluates an... | | `telnyx_conversation_channel` | object | No | The communication channel through w