Code beginner · 3 min read

How to translate text with AI in Python

Direct answer

Use the OpenAI Python SDK to call chat.completions.create with a prompt instructing the model to translate text, specifying the source and target languages.

Setup

Install

bash

pip install openai

Env vars

OPENAI_API_KEY

Imports

python

import os
from openai import OpenAI

Examples

inTranslate 'Hello, how are you?' from English to Spanish.

outHola, ¿cómo estás?

inTranslate 'Good morning, have a nice day!' from English to French.

outBonjour, passez une bonne journée !

inTranslate 'Thank you for your help.' from English to Japanese.

outご助力ありがとうございます。

Integration steps

Import the OpenAI SDK and initialize the client with the API key from os.environ.
Construct a chat message instructing the model to translate the given text specifying source and target languages.
Call the chat.completions.create method with a suitable model like gpt-4o and the messages array.
Extract the translated text from response.choices[0].message.content.
Print or return the translated output.

Full code

python

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

def translate_text(text: str, source_lang: str, target_lang: str) -> str:
    prompt = (
        f"Translate the following text from {source_lang} to {target_lang} without extra explanation:\n" 
        f"{text}"
    )
    messages = [{"role": "user", "content": prompt}]
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=messages
    )
    return response.choices[0].message.content.strip()

if __name__ == "__main__":
    source = "English"
    target = "Spanish"
    text_to_translate = "Hello, how are you?"
    translation = translate_text(text_to_translate, source, target)
    print(f"Original ({source}): {text_to_translate}")
    print(f"Translated ({target}): {translation}")

output

Original (English): Hello, how are you?
Translated (Spanish): Hola, ¿cómo estás?

API trace

Request

json

{"model": "gpt-4o", "messages": [{"role": "user", "content": "Translate the following text from English to Spanish without extra explanation:\nHello, how are you?"}]}

Response

json

{"choices": [{"message": {"content": "Hola, ¿cómo estás?"}}], "usage": {"prompt_tokens": 20, "completion_tokens": 7, "total_tokens": 27}}

Extractresponse.choices[0].message.content

Variants

Streaming translation ›

Use streaming when translating long texts to provide incremental output and improve user experience.

python

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

def translate_stream(text: str, source_lang: str, target_lang: str):
    prompt = f"Translate the following text from {source_lang} to {target_lang} without extra explanation:\n{text}"
    messages = [{"role": "user", "content": prompt}]
    stream = client.chat.completions.create(
        model="gpt-4o",
        messages=messages,
        stream=True
    )
    translation = ""
    for chunk in stream:
        delta = chunk.choices[0].delta.content or ""
        print(delta, end="", flush=True)
        translation += delta
    print()
    return translation.strip()

if __name__ == "__main__":
    translate_stream("Good morning, have a nice day!", "English", "French")

Async translation ›

Use async calls when integrating translation in applications requiring concurrency or non-blocking behavior.

python

import os
import asyncio
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

async def translate_async(text: str, source_lang: str, target_lang: str) -> str:
    prompt = f"Translate the following text from {source_lang} to {target_lang} without extra explanation:\n{text}"
    messages = [{"role": "user", "content": prompt}]
    response = await client.chat.completions.create(
        model="gpt-4o",
        messages=messages
    )
    return response.choices[0].message.content.strip()

async def main():
    translation = await translate_async("Thank you for your help.", "English", "Japanese")
    print(f"Translated (Japanese): {translation}")

if __name__ == "__main__":
    asyncio.run(main())

Use Anthropic Claude for translation ›

Use Anthropic Claude models if you prefer Claude's style or have an Anthropic API key.

python

import os
import anthropic

client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])

system_prompt = "You are a helpful assistant that translates text accurately."

def translate_with_claude(text: str, source_lang: str, target_lang: str) -> str:
    user_message = f"Translate the following text from {source_lang} to {target_lang}: {text}"
    response = client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=1024,
        system=system_prompt,
        messages=[{"role": "user", "content": user_message}]
    )
    return response.content.strip()

if __name__ == "__main__":
    translation = translate_with_claude("Hello, how are you?", "English", "Spanish")
    print(f"Translated (Spanish): {translation}")

Performance

Latency~800ms for gpt-4o non-streaming translation calls

Cost~$0.002 per 500 tokens exchanged on gpt-4o

Rate limitsTier 1: 500 requests per minute / 30,000 tokens per minute

Keep prompts concise to reduce token usage.
Avoid unnecessary system messages or verbose instructions.
Batch multiple sentences in one request to amortize overhead.

Approach	Latency	Cost/call	Best for
Standard OpenAI chat completion	~800ms	~$0.002 per 500 tokens	Simple, accurate translations
Streaming OpenAI chat completion	~800ms + streaming	~$0.002 per 500 tokens	Long texts with incremental output
Anthropic Claude chat completion	~900ms	Check Anthropic pricing	Alternative style and tone preferences

✓

Quick tip

Always specify source and target languages explicitly in your prompt to improve translation accuracy.

⚠

Common mistake

Not specifying source and target languages clearly in the prompt, leading to inaccurate or incomplete translations.

Verified 2026-04 · gpt-4o, claude-3-5-sonnet-20241022

Verify ↗