> ## Documentation Index
> Fetch the complete documentation index at: https://notes.kodekloud.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Project 2 Translating an Article
> Build an article translation tool using OpenAI's chat completion API to translate foreign-language articles into English.
Translate any foreign‐language article into English using OpenAI’s chat completion API. In this tutorial, you’ll:
1. Set up the OpenAI Python client
2. Load the article text into a variable
3. Craft a translation prompt
4. Implement an `article_translator` function
5. Execute the function and display the result
***
## 1. Prerequisites
* Python 3.7+
* An [OpenAI API key](https://platform.openai.com/account/api-keys)
* The `openai` Python package installed:
```bash theme={null}
pip install openai
```
Never hard-code your API key in source files. Use environment variables or a secrets manager:
```bash theme={null}
export OPENAI_API_KEY="sk-..."
```
***
## 2. Initialize the OpenAI Client
Import and configure the client with your API key.
```python theme={null}
import os
from openai import OpenAI
client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
```
***
## 3. Prepare the Article Text & Prompt
Chat models cannot fetch URLs directly, so load your article content into a string. Later, you can extend this for web scraping or file input.
```python theme={null}
# Sample article text in Japanese
article = (
"監視機器大手オリンパスの前社長CEC【高橋雄資】は、警視庁は、"
"自然なカメラマンの金賞受賞者として、東京の都庁と同社の営業法違反の疑いで通報し、"
"8日発表した。"
)
# Build the translation prompt
prompt = f"Translate the following article into English:\n\n{article}"
```
You can replace the hard-coded `article` variable with any string input or file contents for batch translation.
***
## 4. Define the Translator Function
Create a reusable function that sends the system and user messages to the chat completion endpoint.\
A low `temperature` (e.g., 0.1) ensures a faithful translation.
```python theme={null}
def article_translator(prompt: str) -> str:
response = client.chat.completions.create(
model="gpt-4",
messages=[
{
"role": "system",
"content": (
"You are a professional translator specializing in news articles. "
"Provide a direct, accurate English translation without commentary."
),
},
{"role": "user", "content": prompt},
],
temperature=0.1,
)
return response.choices[0].message.content
```
### API Parameters Overview
| Parameter | Description | Example |
| ------------- | --------------------------------------- | ------------- |
| `model` | The chat model to use | `"gpt-4"` |
| `messages` | Conversation history with roles | List of dicts |
| `temperature` | Controls randomness (0 = deterministic) | `0.1` |
***
## 5. Run the Translator
Execute the script and print the translated text.
```python theme={null}
if __name__ == "__main__":
translation = article_translator(prompt)
print(translation)
```
```plaintext theme={null}
$ python translate_article.py
The Tokyo Metropolitan Police Department announced on the 8th that they reported Yusuke Takahashi,
former president of major surveillance-equipment manufacturer Olympus, on suspicion of violating commercial
law by acting as an unauthorized cameraman prize judge in Tokyo’s Metropolitan Government Building.
The department says their investigation will continue carefully.
```
***
## What's Next?
* Batch-translate multiple articles
* Scrape article URLs automatically before translation
* Summarize translated content or detect sentiment
## Links and References
* [OpenAI Python SDK](https://github.com/openai/openai-python)
* [Chat Completion API Docs](https://platform.openai.com/docs/api-reference/chat)
* [Environment Variables Best Practices](https://12factor.net/config)