A flexible FREE and UNLIMITED tool to translate between different languages in a simple way using multiple translators.
Table of Contents
I needed to translate a text using python. It was hard to find a simple way to do it. There are other libraries that can be used for this task, but most of them are buggy, not free, limited, not supported anymore or complex to use.
Therefore, I decided to build this simple tool. It is 100% free, unlimited, easy to use and provide support for all languages.
Basically, my goal was to integrate support for multiple famous translators in this tool.
Install the stable release:
$ pip install -U deep-translator
take a look at the docs if you want to install from source.
from deep_translator import GoogleTranslator
translated = GoogleTranslator(source='auto', target='de').translate("keep it up, you are awesome") # output -> Weiter so, du bist groartig
or using proxies:
from deep_translator import GoogleTranslator
proxies_example = {
"https": "34.195.196.27:8080",
"http": "34.195.196.27:8080"
}
translated = GoogleTranslator(source='auto', target='de', proxies=proxies_example).translate("keep it up, you are awesome") # output -> Weiter so, du bist groartig
or even directly from terminal:
$ deep-translator --source "en" --target "de" --text "hello world"
or shorter
$ dt -tg de -txt "hello world"
In this section, demos on how to use all different integrated translators in this tool are provided.
Note
You can always pass the languages by the name or by abbreviation.
Example: If you want to use english as a source or target language, you can pass english or en as an argument
from deep_translator import (GoogleTranslator,
MicrosoftTranslator,
PonsTranslator,
LingueeTranslator,
MyMemoryTranslator,
YandexTranslator,
PapagoTranslator,
DeeplTranslator,
QcriTranslator,
single_detection,
batch_detection)
Note
You can check the supported languages of each translator by calling the get_supported_languages function.
# default return type is a list
langs_list = GoogleTranslator().get_supported_languages() # output: [arabic, french, english etc...]
# alternatively, you can the dictionary containing languages mapped to their abbreviation
langs_dict = GoogleTranslator().get_supported_languages(as_dict=True) # output: {arabic: ar, french: fr, english:en etc...}
Note
You can also detect language automatically. Notice that this package is free and my goal is to keep it free. Therefore, you will need to get your own api_key if you want to use the language detection function. I figured out you can get one for free here: https://detectlanguage.com/documentation
lang = single_detection('bonjour la vie', api_key='your_api_key')
print(lang) # output: fr
lang = batch_detection(['bonjour la vie', 'hello world'], api_key='your_api_key')
print(lang) # output: [fr, en]
text = 'happy coding'
translated = GoogleTranslator(source='auto', target='de').translate(text=text)
translated = GoogleTranslator(source='auto', target='german').translate(text=text)
# Alternatively, you can pass languages by their abbreviation:
translated = GoogleTranslator(source='en', target='de').translate(text=text)
(Notice that this is important for performance too since instantiating new objects is expensive)
# let's say first you need to translate from auto to german
my_translator = GoogleTranslator(source='auto', target='german')
result = my_translator.translate(text=text)
print(f"Translation using source = {my_translator.source} and target = {my_translator.target} -> {result}")
# let's say later you want to reuse the class but your target is french now
# This is the best practice and how you should use deep-translator.
# Please don't over-instantiate translator objects without a good reason, otherwise you will run into performance issues
my_translator.target = 'fr' # this will override the target 'german' passed previously
result = my_translator.translate(text=text)
print(f"Translation using source = {my_translator.source} and target = {my_translator.target} -> {result}")
# you can also update the source language as well
my_translator.source = 'en' # this will override the source 'auto' passed previously
result = my_translator.translate(text=text)
print(f"Translation using source = {my_translator.source} and target = {my_translator.target} -> {result}")
texts = ["hallo welt", "guten morgen"]
# the translate_sentences function is deprecated, use the translate_batch function instead
translated = GoogleTranslator('de', 'en').translate_batch(texts)
translated = GoogleTranslator(source='auto', target='german').translate_file('path/to/file')
Note
As in google translate, you can use the automatic language detection with mymemory by using "auto" as an argument for the source language. However, this feature in the mymemory translator is not so powerful as in google translate.
text = 'Keep it up. You are awesome'
translated = MyMemoryTranslator(source='auto', target='french').translate(text)
texts = ["hallo welt", "guten morgen"]
# the translate_sentences function is deprecated, use the translate_batch function instead
translated = MyMemoryTranslator('de', 'en').translate_batch(texts)
path = "your_file.txt"
translated = MyMemoryTranslator(source='en', target='fr').translate_file(path)
Note
In order to use the DeeplTranslator translator, you need to generate an api key. Deepl offers a Pro and a free API. deep-translator supports both Pro and free APIs. Just check the examples below. Visit https://www.deepl.com/en/docs-api/ for more information on how to generate your Deepl api key
text = 'Keep it up. You are awesome'
translated = DeeplTranslator(api_key="your_api_key", source="en", target="en", use_free_api=True).translate(text)
Note
deep-translator uses free deepl api by default. If you have the pro version then simply set the use_free_api to false.
texts = ["hallo welt", "guten morgen"]
# the translate_sentences function is deprecated, use the translate_batch function instead
translated = DeeplTranslator("your_api_key").translate_batch(texts)
Note
In order to use the QcriTranslator translator, you need to generate a free api key. Visit https://mt.qcri.org/api/ for more information
# as a property
print("language pairs: ", QcriTranslator("your_api_key").languages)
# as a property
print("domains: ", QcriTranslator("your_api_key").domains)
text = 'Education is great'
translated = QcriTranslator("your_api_key").translate(source='en', target='ar', domain="news", text=text)
# output ->
# see docs for batch translation and more.
word = 'good'
translated_word = LingueeTranslator(source='english', target='french').translate(word)
# pass language by their abbreviation
translated_word = LingueeTranslator(source='en', target='fr').translate(word)
# set the argument return_all to True if you want to get all synonyms of the word to translate
translated_word = LingueeTranslator(source='english', target='french').translate(word, return_all=True)
translated_words = LingueeTranslator(source='english', target='french').translate_words(["good", "awesome"])
Note
You can pass the languages by the name or by abbreviation just like previous examples using GoogleTranslate
word = 'awesome'
translated_word = PonsTranslator(source='english', target='french').translate(word)
# pass language by their abbreviation
translated_word = PonsTranslator(source='en', target='fr').translate(word)
# set the argument return_all to True if you want to get all synonyms of the word to translate
translated_word = PonsTranslator(source='english', target='french').translate(word, return_all=True)
translated_words = LingueeTranslator(source='english', target='french').translate_words(["good", "awesome"])
Note
You need to require an private api key if you want to use the yandex translator. visit the official website for more information about how to get one
lang = YandexTranslator('your_api_key').detect('Hallo, Welt')
print(f"language detected: {lang}") # output -> language detected: 'de'
# with auto detection | meaning provide only the target language and let yandex detect the source
translated = YandexTranslator('your_api_key').translate(source="auto", target="en", text='Hallo, Welt')
print(f"translated text: {translated}") # output -> translated text: Hello world
# provide source and target language explicitly
translated = YandexTranslator('your_api_key').translate(source="de", target="en", text='Hallo, Welt')
print(f"translated text: {translated}") # output -> translated text: Hello world
translated = YandexTranslator('your_api_key').translate_file(source="auto", target="en", path="path_to_your_file")
translated = YandexTranslator('your_api_key').translate_batch(source="auto", target="de", batch=["hello world", "happy coding"])
Note
You need to require an api key if you want to use the microsoft translator. visit the official website for more information about how to get one. Microsoft offers a free tier 0 subscription (2 million characters per month).
Required and optional attributes
There are two required attributes, namely "api_key" (string) and "target" (string or list). Attribute "source" is optional. Also, Microsoft API accepts a number of other optional attributes, you can find them here: https://docs.microsoft.com/azure/cognitive-services/translator/reference/v3-0-translate You can simply add them after the required attributes, see the example.
text = 'happy coding'
translated = MicrosoftTranslator(api_key='some-key', target='de').translate(text=text)
translated_two_targets = MicrosoftTranslator(api_key='some-key', target=['de', 'ru']).translate(text=text)
translated_with_optional_attr = MicrosoftTranslator(api_key='some-key', target='de', textType='html']).translate(text=text)
translated = MicrosoftTranslator(api_key='some-key', target='german').translate(text=text)
# Alternatively, you can pass languages by their abbreviation:
translated = MicrosoftTranslator(api_key='some-key', target='de').translate(text=text)
texts = ["hallo welt", "guten morgen"]
translated = MicrosoftTranslator(api_key='some-key', target='english').translate_batch(texts)
translated = MicrosoftTranslator(api_key='some-key', target='german').translate_file('path/to/file')
Note
You need to require a client id and client secret key if you want to use the papago translator. visit the official website for more information about how to get one.
text = 'happy coding'
translated = PapagoTranslator(client_id='your_client_id', secret_key='your_secret_key', source='en', target='ko').translate(text=text) # output:
Note
Libre translate has multiple mirrors which can be used for the API endpoint. Some require an API key to be used. By default the base url is set to libretranslate.de . This can be set using the "base_url" input parameter.
text = 'laufen'
translated = LibreTranslator(source='auto', target='en', base_url = 'https://libretranslate.com/', api_key = 'your_api_key').translate(text=text) # output: run
translated = LibreTranslator(source='german', target='english').translate(text=text)
# Alternatively, you can pass languages by their abbreviation:
translated = LibreTranslator(source='de', target='en').translate(text=text)
texts = ["hallo welt", "guten morgen"]
translated = LibreTranslator(source='auto', target='en').translate_batch(texts)
translated = LibreTranslator(source='auto', target='en').translate_file('path/to/file')
deep-translator provide out of the box usage of proxies. Just define your proxies config as a dictionary and pass it to the corresponding translator. Below is an example using the GoogleTranslator but this feature can be used with all supported translators.
from deep_translator import GoogleTranslator
# define your proxy configs:
proxies_example = {
"https": "your https proxy", # example: 34.195.196.27:8080
"http": "your http proxy if available"
}
translated = GoogleTranslator(source='auto', target='de', proxies=proxies_example).translate("this package is awesome")
Deep-translator supports a series of command line arguments for quick and simple access to the translators directly in your console.
Note
The program accepts deep-translator
or dt
as a command, feel free to substitute whichever you prefer.
For a list of available translators:
$ deep-translator list
To translate a string or line of text:
$ deep_translator google --source "english" --target "german" --text "happy coding"
Alternate short option names, along with using language abbreviations:
$ deep_translator google -src "en" -tgt "de" -txt "happy coding"
Finally, to retrieve a list of available languages for a given translator:
$ deep-translator languages google
Developers can install the development version of deep-translator and execute unit tests to verify functionality. For more information on doing this, see the contribution guidelines
Check this article on medium to know why you should use the deep-translator package and how to translate text using python. https://medium.com/@nidhalbacc/how-to-translate-text-with-python-9d203139dcf5
If you are facing any problems, please feel free to open an issue. Additionally, you can make contact with the author for further information/questions.
Do you like deep-translator? You can always help the development of this project by:
Take a look in the examples folder for more :) Contributions are always welcome. Read the Contribution guidelines Here
Many thanks to @KirillSklyarenko for his work on integrating the microsoft translator
MIT license
Copyright (c) 2020-present, Nidhal Baccouri
deep-translator offers an api server for easy integration with other applications. Non python applications can communicate with the api directly and leverage the features of deep-translator
Access the api here: https://deep-translator-api.azurewebsites.net/docs
You can download and try the app on play store https://play.google.com/store/apps/details?id=org.translator.translator&hl=en_US&gl=US
After developing the deep-translator, I realized how cool this would be if I can use it as an app on my mobile phone. Sure, there is google translate, pons and linguee apps etc.. but isn't it cooler to make an app where all these translators are integrated?
Long story short, I started working on the app. I decided to use the kivy framework since I wanted to code in python and to develop a cross platform app. I open sourced the Translator++ app on my github too. Feel free to take a look at the code or make a pull request ;)
Note
The Translator++ app is based on the deep-translator package. I just built the app to prove the capabilities of the deep-translator package ;)
I published the first release on google play store on 02-08-2020
Here are some screenshots:
Currently, there are propositions for a website and/or dekstop app based on deep-translator. You can follow the issue here: https://github.com/nidhaloff/deep-translator/issues/144