Skip to main content

Shisa Translation

Shisa Translation is a Japanese-English translation API built on Shisa AI's proprietary large language models. Rather than the literal, phrase-by-phrase output of generic translation tools, it understands context, idioms, and cultural nuance to produce natural, accurate translations in both directions.

POST https://api.shisa.ai/translate/

Send the text and a source/target language pair as multipart/form-data and receive an OpenAI-style response with the translated text at choices[0].message.content. Requests default to the shisa-ai/chotto routing alias; the response model may report the concrete backend model that served the request. Translations can be streamed token-by-token for real-time use.

note

Translation authenticates with a standard bearer token — Authorization: Bearer YOUR_API_KEY, the same header every Shisa service uses. See Authentication for details.

Why Shisa Translation​

  • Context-aware translation — the underlying LLM understands context, idioms, and cultural nuances for natural translations rather than literal word-for-word output.
  • Enterprise security — see the Privacy Policy on the Shisa platform for how your data is handled.
  • Glossary control — preserve up to 20 product names and domain terms with repeated keywords fields.
  • Conversation context — provide prior text with context to improve ambiguous or conversational translations.
  • Real-time translation — stream translations for chat, live content, and interactive applications.

Use cases​

  • Content localization — translate marketing materials, product descriptions, and website content while maintaining brand voice and cultural relevance, across e-commerce product pages, marketing campaigns, blog and SEO content, and mobile app interfaces.
  • Customer support — enable multilingual customer service with real-time translation for live chat, email support, knowledge base articles, and FAQ translations.
  • Documentation & training — translate technical documentation, user manuals, training materials, and internal communications accurately.

Next steps​

  • Quickstart — make your first translation with curl, Python, or JavaScript.
  • API reference — form fields, response fields, and the streaming format.
  • Pricing — how translation usage is billed.