NLLB-200 Website Translator

Translate any website into 200+ languages using Meta's open-source NLLB-200 (No Language Left Behind) AI model via Hugging Face Inference API.

Simply provide URLs and your source/target languages — the actor scrapes the page content, sends it to the NLLB-200 model in chunks, and returns the full translation in a structured dataset.

---

What data does it extract?

Field	Description	Example
url	Page URL	https://example.com/about
title	Page title	"About us — Acme Corp"
sourceLang	Source language code	eng_Latn
targetLang	Target language code	spa_Latn
modelId	NLLB model used	facebook/nllb-200-distilled-600M
originalText	Extracted page text	"Welcome to our company..."
translatedText	Translated text	"Bienvenido a nuestra empresa..."
chunkCount	Number of text chunks translated	4
translatedAt	Timestamp	"2026-04-21T12:00:00Z"

---

Use cases

• International expansion: Translate your competitor's site to understand their value proposition
• Market research: Read foreign-language e-commerce sites, news portals, or government sources
• Content localization: Translate documentation, landing pages, or blogs at scale
• SEO research: Analyze foreign-language keyword content
• Academic research: Access non-English sources at scale

---

How to use

1. Get a free Hugging Face token at huggingface.co/settings/tokens (enable "Inference" permission)
2. Open the actor and configure:
   • URLs to translate: List of pages to process
   • Source language: NLLB code of the original (e.g. eng_Latn for English)
   • Target language: NLLB code for the output (e.g. spa_Latn for Spanish)
   • Hugging Face Token: Your HF API token
   • Crawl depth: 0 for only the given URLs, 1 to also follow internal links
   • Model: Choose speed vs accuracy
3. Download results as JSON or CSV

---

Language codes (most common)

Language	Code
English	eng_Latn
Spanish	spa_Latn
Portuguese	por_Latn
French	fra_Latn
German	deu_Latn
Italian	ita_Latn
Chinese (Simplified)	zho_Hans
Japanese	jpn_Jpan
Korean	kor_Hang
Arabic	arb_Arab
Russian	rus_Cyrl
Hindi	hin_Deva

Full list of 200 language codes: flores200 README

---

Example output (JSON)

{
  "url": "https://example.com/about",
  "title": "About us — Acme Corp",
  "sourceLang": "eng_Latn",
  "targetLang": "spa_Latn",
  "modelId": "facebook/nllb-200-distilled-600M",
  "originalText": "Welcome to Acme Corp. We build great products for...",
  "translatedText": "Bienvenido a Acme Corp. Construimos grandes productos para...",
  "chunkCount": 3,
  "translatedAt": "2026-04-21T12:00:00.000Z"
}

---

Model comparison

Model	Speed	Accuracy	Best for
nllb-200-distilled-600M	⚡ Fast	Good	Most use cases
nllb-200-1.3B	Medium	Better	Balanced
nllb-200-3.3B	Slow	Best	High quality output

---

Pricing

This actor charges $0.002 USD per page translated. A run translating 50 pages costs approximately $0.10 USD, plus your Hugging Face API usage (free tier available).

---

Keywords

website translator, NLLB-200 scraper, multilingual translation actor, Meta AI translation, Hugging Face NLLB, website content translation, 200 languages translator, web page translator API, automatic website translation

---

Legal Disclaimer

This actor is designed to process publicly available content only, in compliance with Chilean Law 19.628 on the Protection of Private Life (Ley 19.628 sobre Protección de la Vida Privada) and general data protection principles.

What this actor does NOT collect:

• Names of natural persons or private individuals
• Personal identification numbers, private emails, or phone numbers
• Any data that could identify or profile a specific private individual

What this actor does:

• Extracts publicly accessible text from web pages (identical to what any browser user would see)
• Sends that text to the Hugging Face Inference API for translation
• Returns the translated content to the user who initiated the run

Users are solely responsible for ensuring their use of this actor and the translated content complies with applicable laws, the terms of service of the websites they scrape, and any relevant copyright regulations in their jurisdiction.