General Text Dialogue Interface Document

1. Introduction

This interface generates dialogue answers by calling OpenAI compatible large language models, supports preset system roles to interact with user questions, and returns the Chinese reply content generated by the model. The core function is to generate responses from the model that conform to the system role settings based on user input questions.

2. request parameters

parameter name	type	necessary	Instructions	example value
api_key	string	yes	Access API key for authentication purposes	sk - XyLy**************************mIqSt
base_url	string	yes	The basic address of the API service, specifying the access path for the model service	https://llm.ai-nebula.com
model	string	yes	Large language model identification for generating answers	claude - sonnet - 4 - 20250514 GPT-4o
messages	array	yew	Dialogue message list, including system role definitions and user questions Each element is an object, including: - role：role(user/system/assistant) - content：message content	{ "role": "user", "content": "hello" }
temperature	number	No	Control the randomness of generated content, with higher 0-2 values indicating stronger response randomness	0.7

3. Return Value Description

The function returns a result of string type, which includes two possibilities:

(1) Successful Response

Type: String

Description: The Chinese response content generated by the model

Example:

Artificial intelligence is a new technical science that researches and develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence. It attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a manner similar to human intelligence.

(2) Error Response

Type: String

Description: A description containing the cause of the error. Possible situations:

Error: API key is invalid or unauthorized (key verification failed)

Error: Model claude-sonnet-4-20250514 does not exist or is not supported (model does not exist)

Error: Unable to connect to the API server (network or address error)

API request failed: [specific error information] (other API layer errors)

Unknown error occurred: [specific error information] (uncaught exception)

4. Error Handling

The function captures and handles the following exception types (based on the OpenAI Python SDK exception system):

Exception Type	Trigger Scenario	Return Message
AuthenticationError	API key is invalid, expired, or unauthorized	Error: API key is invalid or unauthorized
NotFoundError	The specified model ID does not exist or is not supported by the server	Error: Model [model name] does not exist or is not supported
APIConnectionError	Network interruption, base_url error, or server unresponsive	Error: Unable to connect to the API server
APIError	Server-side exceptions such as API request format errors, insufficient permissions, etc.	API request failed: [error details]
Other Uncaught Exceptions	Code logic errors, unexpected data formats, etc.	Unknown error occurred: [error details]

5. Implementation Code

6. Usage Example Output

(1) Successful Output Example

AI Response Content:
Artificial intelligence is a new technical science that researches and develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence. Its core includes fields such as machine learning, natural language processing, and computer vision. It is widely applied in industries like healthcare, finance, and transportation, and is profoundly changing human production and lifestyles.

(2) Error Output Example

AI Response Content:
Error: API key is invalid or unauthorized

7. Important Notes

Parameters such as api_key, base_url, and model in the code are required values. If modifications are needed, please adjust the corresponding fields directly within the function.

The messages list contains conversation history; the system role is used to set model behavior, and the user role represents the current user question.

It is recommended to set the temperature parameter between 0 and 2, where 0 indicates more deterministic generated content, and 2 indicates greater randomness.

Network stability significantly affects the success rate of API calls; it is recommended to use this function in a good network environment.

To extend to multi-turn conversation, modify the messages parameter by appending historical conversation records (including responses from the assistant role).

The code depends on the openai Python library; please ensure it is installed before use (pip install openai).

General Text Dialogue Interface Document

1. Introduction#

2. request parameters#

3. Return Value Description#

(1) Successful Response#

(2) Error Response#

4. Error Handling#

5. Implementation Code#

6. Usage Example Output#

(1) Successful Output Example#

(2) Error Output Example#

7. Important Notes#