Real-Time Text Translation

​

Authentication

​

Request body

BASE_URL='https://api-integration.ollang.com'

curl --no-buffer --fail-with-body \
  "${BASE_URL}/integration/real-time-text-translation" \
  -H "X-Api-Key: ${OLLANG_API_KEY}" \
  -H 'Content-Type: application/json' \
  -d '{
    "text": "Hello, how are you?",
    "targetLanguage": "Turkish",
    "sourceLanguage": "English"
  }'

​

Streaming response

data: {"choices":[{"delta":{"content":"Merhaba, "}}]}

data: {"choices":[{"delta":{"content":"nasılsın?"}}]}

data: {"choices":[{"delta":{},"finish_reason":"stop"}]}

data: [DONE]

​

Reading the stream

• Append each choices[0].delta.content string to assemble the translation. Events without text, including role, finish, and usage events, are not translation text.
• Buffer incomplete SSE events across network reads. A network chunk is not necessarily one complete event or one complete UTF-8 character.
• [DONE] marks the end of the stream. Usage events may precede it and report tokens, not Ollang credit deductions.
• Inspect finish_reason; a length finish indicates truncated output.
• A disconnect cancels the translation. If the stream ends without [DONE], treat it as incomplete. Errors after streaming starts may terminate the connection rather than return a new HTTP status.

​

Preview limits and errors