Google Translate

Google Translate compared with DeepL, and Microsoft Translator
Google Translate	DeepL	Microsoft Translator
Language coverage: The web interface lists roughly 249 languages, the widest list of the three.	Language coverage: The vendor table lists 114 languages for text translation. Formal or informal tone and glossaries are not available for every language.	Language coverage: Azure AI Translator states more than 100 supported languages and dialects, and a Languages operation returns the current list.
Inputs: Text with language detection and voice input, documents (.docx, .pdf, .pptx, .xlsx), websites by URL and images (.jpg, .jpeg, .png, .webp).	Inputs: Documents in PDF, Word, PowerPoint and Excel formats, with formatting preserved according to the vendor. Apps exist for Windows, Mac, Android, iPhone and iPad.	Inputs: Text through a REST API or SDKs, with Translate, Transliterate and Languages operations. Transliteration converts script, which suits non-Latin names.
API and cost: Cloud Translation Basic is paid per character after a free credit each month covering the first 500,000 characters. Advanced adds LLM translation and custom models, with document translation billed per page.	API and cost: Free use is limited to 50,000 characters and 1 file monthly. The Individual plan is a paid subscription billed annually.	API and cost: A free tier of 2 million characters monthly is listed. Pay as you go is billed per million characters, with commitment tiers of 250 million, 1 billion or 4 billion characters, and LLM translation billed per token.
Terminology control: Cloud glossary creation is not charged. Custom AutoML models can be trained for a fee.	Terminology control: Glossaries are limited to 1 on Individual, 5 on Team and unlimited on Business, which also adds translation memory.	Terminology control: Adaptive custom translation accepts up to five reference translations to steer LLM output. Tone and gender controls apply to LLM translation.
	Deployment and data: Pro subscriber content is stored only as technically needed and is never used to train models. The vendor holds ISO 27001 and SOC 2 Type II certifications.	Deployment and data: A Docker container is available for disconnected or controlled scenarios. The endpoint chosen determines where requests are processed.

Google Translate compared with DeepL, and Microsoft Translator

Google Translate

DeepL

Microsoft Translator

Language coverage: The web interface lists roughly 249 languages, the widest list of the three.

Language coverage: The vendor table lists 114 languages for text translation. Formal or informal tone and glossaries are not available for every language.

Language coverage: Azure AI Translator states more than 100 supported languages and dialects, and a Languages operation returns the current list.

Inputs: Text with language detection and voice input, documents (.docx, .pdf, .pptx, .xlsx), websites by URL and images (.jpg, .jpeg, .png, .webp).

Inputs: Documents in PDF, Word, PowerPoint and Excel formats, with formatting preserved according to the vendor. Apps exist for Windows, Mac, Android, iPhone and iPad.

Inputs: Text through a REST API or SDKs, with Translate, Transliterate and Languages operations. Transliteration converts script, which suits non-Latin names.

API and cost: Cloud Translation Basic is paid per character after a free credit each month covering the first 500,000 characters. Advanced adds LLM translation and custom models, with document translation billed per page.

API and cost: Free use is limited to 50,000 characters and 1 file monthly. The Individual plan is a paid subscription billed annually.

API and cost: A free tier of 2 million characters monthly is listed. Pay as you go is billed per million characters, with commitment tiers of 250 million, 1 billion or 4 billion characters, and LLM translation billed per token.

Terminology control: Cloud glossary creation is not charged. Custom AutoML models can be trained for a fee.

Terminology control: Glossaries are limited to 1 on Individual, 5 on Team and unlimited on Business, which also adds translation memory.

Terminology control: Adaptive custom translation accepts up to five reference translations to steer LLM output. Tone and gender controls apply to LLM translation.

Deployment and data: Pro subscriber content is stored only as technically needed and is never used to train models. The vendor holds ISO 27001 and SOC 2 Type II certifications.

Deployment and data: A Docker container is available for disconnected or controlled scenarios. The endpoint chosen determines where requests are processed.