A Family of Text and Translation Models
Mayura is built for real-world Indian language tasks, including translation, transliteration, and language identification across Indian Languages
Say hello to Mayura
Announcing the launch of best-in-class, formal & colloquial translation model India’s top 10 languages
Making information accessible for India
.svg)
A Family of Text and Translation Models
Translate like you talk
Mayura’s translation models convert text between English and Indian languages while preserving meaning, context, and tone. Unlike typical systems that rely heavily on English as an intermediary, Mayura also supports direct Indic-to-Indic translation, allowing for more accurate and fluent results.
Transliterate Text while preserving pronunciation
Transliteration converts text from one script to another while preserving pronunciation. It's especially useful for users who speak an Indian language but prefer typing in Roman script. The model also handles numeric conversions between Indic and Latin digits, which is essential for tasks like text-to-speech and form inputs.
Identify language and script accurately
Identify the language and script of any input text — even when it's short, code-mixed, or ambiguous. This capability is essential in multilingual systems where the input language isn't known in advance. By detecting the correct language and script early, it enables precise routing to the right translation or transliteration model.
Pricing
Translate
₹ 20 / 10k characters
Transliteration
₹ 20 / 10k characters
Language Identification
₹ 3.5 / 10k characters
