AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

azure-ai-translation-ts

Build translation applications using Azure Translation SDKs for JavaScript (@azure-rest/ai-translation-text, @azure-rest/ai-translation-document). Use when implementing text translation, transliteration, language detection, or batch document translation.

Original source
microsoft
License
MIT
Pinned revision
3f5b461b2314e01c400564051e92469114c5b68f

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

Ask your assistant

Use Agentic Memory skills to read microsoft--azure-ai-translation-ts at revision 3f5b461b2314e01c400564051e92469114c5b68f, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 1

SKILL.md

Azure Translation SDKs for TypeScript

Text and document translation with REST-style clients.

Installation

# Text translation
npm install @azure-rest/ai-translation-text @azure/identity

# Document translation
npm install @azure-rest/ai-translation-document @azure/identity

Environment Variables

TRANSLATOR_ENDPOINT=https://api.cognitive.microsofttranslator.com
TRANSLATOR_SUBSCRIPTION_KEY=<your-api-key>
TRANSLATOR_REGION=<your-region>  # e.g., westus, eastus
AZURE_TOKEN_CREDENTIALS=prod # Required only if DefaultAzureCredential is used in production

Text Translation Client

Authentication

import TextTranslationClient, { TranslatorCredential } from "@azure-rest/ai-translation-text";

// API Key + Region
const credential: TranslatorCredential = {
  key: process.env.TRANSLATOR_SUBSCRIPTION_KEY!,
  region: process.env.TRANSLATOR_REGION!,
};
const client = TextTranslationClient(process.env.TRANSLATOR_ENDPOINT!, credential);

// Or just credential (uses global endpoint)
const client2 = TextTranslationClient(credential);

Translate Text

import TextTranslationClient, { isUnexpected } from "@azure-rest/ai-translation-text";

const response = await client.path("/translate").post({
  body: {
    inputs: [
      {
        text: "Hello, how are you?",
        language: "en",  // source (optional, auto-detect)
        targets: [
          { language: "es" },
          { language: "fr" },
        ],
      },
    ],
  },
});

if (isUnexpected(response)) {
  throw response.body.error;
}

for (const result of response.body.value) {
  for (const translation of result.translations) {
    console.log(`${translation.language}: ${translation.text}`);
  }
}

Translate with Options

const response = await client.path("/translate").post({
  body: {
    inputs: [
      {
        text: "Hello world",
        language: "en",
        textType: "Plain",  // or "Html"
        targets: [
          {
            language: "de",
            profanityAction: "NoAction",  // "Marked" | "Deleted"
            tone: "formal",  // LLM-specific
          },
        ],
      },
    ],
  },
});

Get Supported Languages

const response = await client.path("/languages").get();

if (isUnexpected(response)) {
  throw response.body.error;
}

// Translation languages
for (const [code, lang] of Object.entries(response.body.translation || {})) {
  console.log(`${code}: ${lang.name} (${lang.nativeName})`);
}

Transliterate

const response = await client.path("/transliterate").post({
  body: { inputs: [{ text: "这是个测试" }] },
  queryParameters: {
    language: "zh-Hans",
    fromScript: "Hans",
    toScript: "Latn",
  },
});

if (!isUnexpected(response)) {
  for (const t of response.body.value) {
    console.log(`${t.script}: ${t.text}`);  // Latn: zhè shì gè cè shì
  }
}

Detect Language

const response = await client.path("/detect").post({
  body: { inputs: [{ text: "Bonjour le monde" }] },
});

if (!isUnexpected(response)) {
  for (const result of response.body.value) {
    console.log(`Language: ${result.language}, Score: ${result.score}`);
  }
}

Document Translation Client

Authentication

import DocumentTranslationClient from "@azure-rest/ai-translation-document";
import { DefaultAzureCredential, ManagedIdentityCredential } from "@azure/identity";

const endpoint = "https://<translator>.cognitiveservices.azure.com";

// Local dev: DefaultAzureCredential. Production: set AZURE_TOKEN_CREDENTIALS=prod or AZURE_TOKEN_CREDENTIALS=<specific_credential>
const credential = new DefaultAzureCredential({requiredEnvVars: ["AZURE_TOKEN_CREDENTIALS"]});
// Or use a specific credential directly in production:
// See https://learn.microsoft.com/javascript/api/overview/azure/identity-readme?view=azure-node-latest#credential-classes
// const credential = new ManagedIdentityCredential();

// TokenCredential
const client = DocumentTranslationClient(endpoint, credential);

// API Key
const client2 = DocumentTranslationClient(endpoint, { key: "<api-key>" });

Single Document Translation

import DocumentTranslationClient from "@azure-rest/ai-translation-document";
import { writeFile } from "node:fs/promises";

const response = await client.path("/document:translate").post({
  queryParameters: {
    targetLanguage: "es",
    sourceLanguage: "en",  // optional
  },
  contentType: "multipart/form-data",
  body: [
    {
      name: "document",
      body: "Hello, this is a test document.",
      filename: "test.txt",
      contentType: "text/plain",
    },
  ],
}).asNodeStream();

if (response.status === "200") {
  await writeFile("translated.txt", response.body);
}

Batch Document Translation

import { ContainerSASPermissions, BlobServiceClient } from "@azure/storage-blob";

// Generate SAS URLs for source and target containers
const sourceSas = await sourceContainer.generateSasUrl({
  permissions: ContainerSASPermissions.parse("rl"),
  expiresOn: new Date(Date.now() + 24 * 60 * 60 * 1000),
});

const targetSas = await targetContainer.generateSasUrl({
  permissions: ContainerSASPermissions.parse("rwl"),
  expiresOn: new Date(Date.now() + 24 * 60 * 60 * 1000),
});

// Start batch translation
const response = await client.path("/document/batches").post({
  body: {
    inputs: [
      {
        source: { sourceUrl: sourceSas },
        targets: [
          { targetUrl: targetSas, language: "fr" },
        ],
      },
    ],
  },
});

// Get operation ID from header
const operationId = new URL(response.headers["operation-location"])
  .pathname.split("/").pop();

Get Translation Status

import { isUnexpected, paginate } from "@azure-rest/ai-translation-document";

const statusResponse = await client.path("/document/batches/{id}", operationId).get();

if (!isUnexpected(statusResponse)) {
  const status = statusResponse.body;
  console.log(`Status: ${status.status}`);
  console.log(`Total: ${status.summary.total}`);
  console.log(`Success: ${status.summary.success}`);
}

// List documents with pagination
const docsResponse = await client.path("/document/batches/{id}/documents", operationId).get();
const documents = paginate(client, docsResponse);

for await (const doc of documents) {
  console.log(`${doc.id}: ${doc.status}`);
}

Get Supported Formats

const response = await client.path("/document/formats").get();

if (!isUnexpected(response)) {
  for (const format of response.body.value) {
    console.log(`${format.format}: ${format.fileExtensions.join(", ")}`);
  }
}

Key Types

// Text Translation
import type {
  TranslatorCredential,
  TranslatorTokenCredential,
} from "@azure-rest/ai-translation-text";

// Document Translation
import type {
  DocumentTranslateParameters,
  StartTranslationDetails,
  TranslationStatus,
} from "@azure-rest/ai-translation-document";

Best Practices

  1. Auto-detect source - Omit language parameter to auto-detect
  2. Batch requests - Translate multiple texts in one call for efficiency
  3. Use SAS tokens - For document translation, use time-limited SAS URLs
  4. Handle errors - Always check isUnexpected(response) before accessing body
  5. Regional endpoints - Use regional endpoints for lower latency