Create Translation Stream

This endpoint provides instantaneous text translation with adaptive context handling, streaming results character-by-character as they’re processed. This approach enables real-time subtitling, live chat translation, and dynamic content localization without waiting for full document processing.

​

Customizing Your translation

​

Contextual Adaptation Parameters

Enhance your translations with linguistic context controls:

• Formality Level (Optional):

  Setting	Value	Use Case
  Formal	1	Business documents, academic papers, official correspondence
  Casual	2	Chat messages, social media, casual communication

• Grammatical Gender (Optional):

  Applies to languages with gender-specific grammar

  Setting	Value
  Not Specified	0
  Male	1
  Female	2
  Neutral	9

​

Language Support

• Configure your translation pipeline using:

  • /source-languages - Supported input languages

  • /target-languages - Available output languages

​

Implementation Guide

Streaming Translation Workflow:

import requests

translation_config = {
    "text": "Jupiter, the largest planet in our solar system...",
    "source_language": 1,   # English
    "target_language": 81,  # Hindi
    "formality": 2,         # Formal
    "gender": 1             # Male grammatical gender
}

def sync_translation_stream(config):
    """Handle translation stream synchronously"""
    headers = {"x-api-key": "your_api_key_here"}

    with requests.post(
        "https://client.camb.ai/apis/translation/stream",
        json=config,
        headers=headers,
        stream=True
    ) as response:

        # Verify successful connection
        response.raise_for_status()

        # Display credit cost
        print(f"Credits required: {response.headers.get('X-Credits-Required', 'Unknown')}")

        # Process stream chunks
        full_translation = ""
        for chunk in response.iter_content(chunk_size=128):
            try:
                decoded = chunk.decode('utf-8')
                full_translation += decoded
                print(decoded, end='', flush=True)
            except UnicodeDecodeError:
                print("�", end='')  # Error placeholder

        return full_translation

# Execute the translation
translated_text = sync_translation_stream(translation_config)
print("\nComplete Translation:\n", translated_text)

​

Response Handling

• Stream Format: UTF-8 encoded text/event-stream with incremental translations.

• Credit Tracking: X-Credits-Required header shows computational resources used.

• Error Handling: Failed streams immediately terminate with error message.

​

Use Case Examples

• Live Caption Translation: Convert spoken language subtitles in real-time

• Chat Localization: Instant message translation for multilingual support teams

• Document Preview: Stream translated content while full processing continues

This streaming solution is ideal for applications requiring immediate partial results while maintaining full translation context awareness. The character-level streaming allows for dynamic UI updates and progressive content rendering.