For more than a decade, Google Translate has established itself as the go-to tool for machine translation, with a simple promise: to translate instantly into as many languages as possible. Today, the tool boasts more than 500 million daily users and processes over 100 billion words a day, becoming a universal go-to1. But the rise of generative language models has profoundly changed expectations. Translation is no longer just about converting words; it is now about conveying meaning, tone, and sometimes even intent. It is in this context that OpenAI appears to be preparing a credible alternative, with a dedicated interface unofficially dubbed ChatGPT Translate.
A dedicated translation interface is already available
Unlike a simple feature built into the chatbot, OpenAI offers a standalone web page dedicated exclusively to translation. This interface is already publicly available, without any prior official announcement, at: https://chatgpt.com/translate.
It is also indexed by Google, which confirms that this is not a closed internal test, but rather a feature that has been launched, likely in a pre-deployment phase2.
Geographically, the tool is accessible in both the United States and Europe, including France, with no apparent location-based restrictions. The interface automatically adapts to the browser’s language, suggesting an international focus right from the start.
More than 50 languages and clearly identified uses
According to the information on the page, ChatGPT Translate supports over 50 languages, covering the main everyday translation needs. OpenAI explicitly targets several user groups, including language learners, travelers, and professionals who work with multilingual content. This segmentation marks a notable difference from Google Translate, which was historically designed as a universal, general-purpose tool.
A contextual translation rather than a literal one
The key difference introduced by ChatGPT Translate lies in the very nature of the underlying model. Whereas Google Translate relies on architectures optimized for linguistic correspondence, ChatGPT uses a generative model capable of understanding context and intent. In practical terms, users can adjust the translation, request a more formal, educational, or natural rephrasing, and refine the result. Translation thus becomes an interactive process, rather than a static result.
This approach responds to a documented shift in user behavior. According to several industry studies, more than 60% of professional users now expect a translation tool to take context and register into account, beyond mere lexical accuracy3.
Google Translate Faces a New Kind of Competition
Google Translate retains significant structural advantages, including very broad language coverage, browser integration, and the ability to translate documents, images, and entire web pages. ChatGPT Translate, at this stage, appears to be more text-focused. However, OpenAI introduces a decisive added value: the ability to explain, adapt, and enrich translations by linking them to other functions such as rephrasing, summarization, or language learning.
This difference illustrates a paradigm shift. Google Translate excels at providing instant, universal translation, while ChatGPT Translate is designed to offer enhanced language assistance.
Challenges and Limitations of Generative Translation
This new approach is not without its limitations. Like any generative AI, ChatGPT Translate relies on probabilistic models that may produce plausible but questionable translations, particularly in sensitive, legal, medical, or regulatory contexts. Recent academic research shows that, while LLMs achieve high levels of quality, human oversight remains essential for high-stakes translations4.
Furthermore, the issue of confidentiality regarding translated texts remains a key concern for professional use, a point on which OpenAI will need to provide clear assurances if the tool is to be adopted by businesses.
Toward a New Approach to Computer-Assisted Translation
With ChatGPT Translate, OpenAI isn’t just offering yet another competitor to Google Translate. The company is outlining a different vision: one of translation integrated into a conversational ecosystem, capable of evolving, explaining itself, and adapting. If this feature were to be officially launched and enhanced, it could permanently transform the way users interact with languages, making translation an intelligent, iterative, and contextualized process.
Learn more
As part of this ongoing transformation of language use driven by artificial intelligence, check out our article “AI and Speech: Voxtral, Mistral’s Open-Source Response to Large Language Models”, which analyzes recent advances in language and speech processing technologies, as well as their implications for multilingual communication.
References
1. SEO Sandwitch. (2025). Google Translate Statistics.
https://seosandwitch.com/google-translate-statistics/
2. OpenAI. (2026). ChatGPT Translate. https://chatgpt.com/translate
3. CSA Research. (2024). The Evolving Expectations of Machine Translation Users.
https://csa-research.com
4. Yan, J. et al. (2024). GPT-4 vs. Human Translators.
https://arxiv.org/abs/2407.03658
