Skip to content
4All API Docs

OpenAI Chat Format (Python)

OpenAI Chat Format (Python)

Section titled “OpenAI Chat Format (Python)”

Overview

Official Documentation

OpenAI Chat

📝 Introduction

Section titled “📝 Introduction”

Given a list of messages containing a conversation, the model will return a response. For related guidance, please refer to the official OpenAI documentation: Chat Completions

💡 Basic Python Request Example: Text Conversation ✅

Section titled “💡 Basic Python Request Example: Text Conversation ✅”

1. Create the .env variable file

Section titled “1. Create the .env variable file”
  • In the same directory as your Python script (for example, xxxx.py), create a file named .env (note: the filename must be exactly .env, with a leading dot and no other prefix).
  • In that .env file, add the following content and fill in your API key and base URL: # .env file content OPENAI_API_KEY=“sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx” OPENAI_BASE_URL=“https://sg.4All API.com/v1”

You can add other configuration variables in the future, for example:

Section titled “You can add other configuration variables in the future, for example:”

DEFAULT_MODEL=“gpt-4.1”

Section titled “DEFAULT_MODEL=“gpt-4.1””

MAX_TOKENS_DEFAULT=150

Section titled “MAX_TOKENS_DEFAULT=150”
### 2. Create a `xxxx.py` file and write the following Python code.
- In the same directory as your Python script (for example, `xxxx.py`), create a file named `.env` (note: the filename must be exactly `.env`, with a leading dot and no other prefix).
```python
import os
from dotenv import load_dotenv, find_dotenv # 导入 find_dotenv 帮助定位
import openai
from openai import OpenAI
import httpx


# 0. Load environment variables from the .env file (enhanced debugging)
print("--- 开始 .env 文件加载调试 ---")
# Try to locate the .env file in the current working directory or the script directory
# find_dotenv(usecwd=True) will try the current working directory first
# find_dotenv() (no arguments) will search upward starting from the script location
dotenv_path_found = find_dotenv(usecwd=True) # Check the current working directory
if not dotenv_path_found:
    dotenv_path_found = find_dotenv() # If not found in the CWD, search normally (from the script directory upward)


if dotenv_path_found:
    print(f"DEBUG: 找到 .env 文件路径: {dotenv_path_found}")
    # verbose=True will print detailed information about the loading process
    # override=True means variables in the .env file will override existing environment variables with the same name
    loaded_successfully = load_dotenv(dotenv_path=dotenv_path_found, verbose=True, override=True)
    if loaded_successfully:
        print("DEBUG: 成功从 .env 文件加载变量。")
    else:
        # If loaded_successfully is False, it may mean the file is empty or could not be parsed,
        # but usually as long as the file is found and not empty, python-dotenv may still return True even if the contents have issues.
        # The real check is whether os.getenv can retrieve the values afterward.
        print("DEBUG: .env 文件已找到，但 load_dotenv() 执行完毕 (请检查 verbose 输出和后续变量值)。")
else:
    print("DEBUG: 未能找到 .env 文件。")
    print("DEBUG: 请确保名为 '.env' 的文件存在于脚本所在目录或项目的根目录中。")


print("--- .env 文件加载调试结束 ---")
print("--- 开始环境变量获取调试 ---")


# 1. Load the API key and base URL from environment variables
api_key = os.getenv("OPENAI_API_KEY")
base_url_from_env = os.getenv("OPENAI_BASE_URL")


# Print the raw values retrieved for debugging
print(f"DEBUG: os.getenv(\"OPENAI_API_KEY\") 返回的值: {'一个字符串 (已隐藏具体内容)' if api_key else 'None'}")
if api_key:
    print(f"DEBUG: API Key 的前5个字符: {api_key[:5]}") # Print part of it for confirmation


print(f"DEBUG: os.getenv(\"OPENAI_BASE_URL\") 返回的值: {base_url_from_env if base_url_from_env else 'None'}")
print("--- 环境变量获取调试结束 ---")


if not api_key:
    print("--------------------------------------------------------------------")
    print("错误：未能从 .env 文件或环境变量中获取 OPENAI_API_KEY。")
    print("请仔细检查以下几点：")
    print("1. 项目根目录或脚本所在目录中是否存在一个名为 '.env' 的文件。")
    print("   (DEBUG 信息中 '找到 .env 文件路径:' 是否显示了正确的路径？)")
    print("2. '.env' 文件中是否正确定义了 OPENAI_API_KEY='your_actual_key'。")
    print("   (确保键名拼写正确，无多余空格，API密钥值完整无误)。")
    print("3. 确保您的 .env 文件内容与示例格式一致，例如：")
    print("   OPENAI_API_KEY=\"sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx\"")
    print("   OPENAI_BASE_URL=\"https://sg.4All API.com/v1\"")
    print("请查看上面以 'DEBUG:' 开头的详细输出，以帮助定位问题。")
    print("--------------------------------------------------------------------")
    exit() # Required API key not found, exit the program
else:
    print(f"成功加载的 API Key (部分显示): '{api_key[:5]}...{api_key[-4:]}'")


# The base URL can be loaded from the environment variables; if not set, use default_base_url = "https://sg.4All API.com/v1"
default_base_url = "OPENAI_BASE_URL" # 您常用的 URL 作为默认值
base_url = base_url_from_env if base_url_from_env else default_base_url
print(f"使用的 Base URL: {base_url}")


if base_url == default_base_url and not base_url_from_env :
    print(f"(提示: OPENAI_BASE_URL 未在 .env 文件或环境变量中指定, 当前使用的是代码中的默认值 '{default_base_url}'。)")


# ... (后续的 OpenAI 客户端初始化、API 调用和错误处理代码保持不变) ...
# 2. Configure the API client
client = OpenAI(
    api_key=api_key,
    base_url=base_url,
    timeout=httpx.Timeout(300.0, connect=60.0),
    max_retries=1,
)


# 3. Prepare the message body for the API request
messages = [
    {"role": "user", "content": "你好，你好，你能做什么？请用中文回答。"}
]


# 4. Send the request and handle the response
try:
    print("正在尝试调用 OpenAI API...")
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=messages,
        max_tokens=4500,
        temperature=0.8,
    )
    if response.choices:
        assistant_reply = response.choices[0].message.content
        print("模型回复:", assistant_reply)
    else:
        print("未能从 API 获取有效回复。")
except openai.AuthenticationError as e:
    print(f"OpenAI API 认证失败: {e}")
    print("这通常意味着 API 密钥无效或没有权限。请再次核对 .env 文件中的 OPENAI_API_KEY 是否为您从 4All API 获取的正确密钥。")
    print(f"当前尝试使用的 API Key (来自 .env 或环境变量，部分显示): '{api_key[:5]}...{api_key[-4:]}'，Base URL 为: {base_url}")
except openai.APIConnectionError as e:
    print(f"无法连接到 OpenAI API: {e}")
except openai.RateLimitError as e:
    print(f"达到 OpenAI API速率限制: {e}")
except openai.APIStatusError as e:
    print(f"OpenAI API 返回了错误状态码: {e.status_code}")
    print(f"响应详情: {e.response}")
except Exception as e:
    print(f"调用 API 时发生未知错误: {e}")
    print(f"错误类型: {type(e).__name__}")

Please:

  • Carefully check your .env file: Filename: make sure it is .env. Location: make sure it is in the same folder as your Python script. Content: confirm again that the key names (OPENAI_API_KEY, OPENAI_BASE_URL) are correct, and that the API key and URL values are also completely correct, with no extra or missing characters. The format should look like this: Code snippet OPENAI_API_KEY= “sk-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx” OPENAI_BASE_URL= “https://sg.4All API.com/v1”
  • Run the complete Python script provided above.
  • Observe the DEBUG: output at the beginning of the script. Does it find the .env file (DEBUG: 找到 .env 文件路径: ...)? Does load_dotenv report that it loaded successfully (DEBUG: 成功从 .env 文件加载变量。)? What does os.getenv("OPENAI_API_KEY") return (DEBUG: os.getenv("OPENAI_API_KEY") 返回的值: ...)?

Based on the DEBUG output:

  • If the DEBUG messages show that the .env file could not be found, or os.getenv("OPENAI_API_KEY") returns None, then the issue lies in the .env file discovery or loading process. Please focus on checking the filename, location, and the related DEBUG output from load_dotenv.
  • If the DEBUG messages show that the .env file was found and loaded, and os.getenv("OPENAI_API_KEY") successfully returns a string that looks like a key, but you still get an AuthenticationError later, then the problem is not the environment-variable loading mechanism. Instead, the value of OPENAI_API_KEY in the .env file itself is invalid or not authorized for the https://sg.4All API.com/v1 service. In that case, you need to contact the provider of 4All API to confirm the validity of the key.