About this tool
Use this page when your source is written in Khmer and you need an English draft for a message, a sign, a study note, or a first look at a document. It opens with Khmer as the source and English as the target, so the direction is ready before you paste. Khmer text often relies on context that is not spelled out in the same way as an English sentence, and word boundaries are not normally marked with a visible space between every word. A translation system has to interpret that continuous script and can choose a different segmentation or meaning than the writer intended. Khmer expressions may also carry a relationship, level of respect, or local idiom that a short English sentence cannot fully preserve. Review who is speaking, who is being addressed, and what happened before the sentence. For scanned text, OCR can mistake stacked characters or diacritics, so correct the extracted Khmer before translating. This page accepts Unicode text; if your source uses Limon encoding, convert it with the Limon-to-Unicode tool first. The result is Google machine translation sent from your browser to Google, not an official or certified translation.
How it works
The page starts with Khmer as the source and English as the target. Paste or type Unicode Khmer, check the direction labels, and press Translate; the browser sends the text to Google’s translate-pa endpoint and displays its machine-generated English. Long input is divided into chunks of up to 1,024 characters, with newline and space boundaries preferred, and the returned pieces are joined. In Khmer, a visible space commonly separates phrases rather than each word, so a cut at an arbitrary character boundary may separate information the system needs to interpret together. You can also select an image and run Tesseract.js OCR in the browser, then inspect and edit the recognised Khmer before sending it for translation. Supported document files are uploaded to InKhmer’s extraction API. If an OCR result contains broken stacks, missing marks, or a name rendered incorrectly, fix the source text first; English output cannot recover letters that were never recognised. For idioms, family terms, names, and formal messages, compare the draft with the original and ask a fluent Khmer speaker to review uncertain parts. The conversion step mentioned above lives at khmer-legacy-to-unicode-converter, which handles Limon, ABC-Zwsp and Khek with auto-detection. The paired khmer-unicode-to-legacy-converter covers the reverse direction for print workflows that still need legacy-encoded text.
Use cases
- Read the gist of a Khmer shop sign before asking a worker to explain a detail.
- Translate a short Khmer chat message into English before drafting a careful response.
- Turn typed Khmer study notes into an English draft for vocabulary comparison.
- Extract Khmer from a clear photo of a menu and check the OCR before translating.
- Understand a Khmer notice while travelling, then confirm dates and instructions locally.
- Prepare an initial English rendering of a Khmer paragraph for a human reviewer.
Frequently asked questions
Why does the English result miss a word or phrase?
Written Khmer does not normally place a visible space between each word; spaces often mark phrase boundaries instead. The system has to infer where units belong, and a different segmentation can change the reading. A short fragment gives it less context, so include the complete sentence or neighbouring phrase when appropriate. Check the original with a fluent Khmer reader when a missed word changes an instruction, amount, or relationship.
Can I paste text from a legacy Khmer font?
This page expects Unicode Khmer text. Some older Khmer documents store characters using a legacy font encoding, which may appear as Latin letters or scrambled glyphs when pasted into a Unicode text box. If your text uses Limon, convert it with the Limon-to-Unicode converter before translating. Other font encodings need their matching conversion method; changing the font display alone does not make the underlying text Unicode.
Can I translate Khmer text from a photo?
Yes. The image upload uses Tesseract.js OCR in your browser with Khmer and English recognition data, then puts the extracted text into the source box. Read and correct that text before pressing Translate. A blurred photo, decorative type, handwritten notes, or marks touching nearby characters can lead to recognition errors that affect the English result. For a scanned document that matters, have a person compare the OCR against the image.
Will an idiom or family term have the same meaning in English?
Not always. A literal rendering may lose the familiar, respectful, teasing, or figurative meaning of a Khmer expression. Family and relationship terms can also depend on how people address one another rather than only on biological relation. Use the English result as a clue, keep the Khmer original beside it, and ask a fluent reader to explain the intended sense when the exact meaning affects your response.
Does this provide a certified translation for official use?
No. This page produces Google machine translation and does not include review by a sworn translator, certification, signature, or legal validation. It is not an official translation and should not be filed as a certified document. Ask the office or organisation receiving your document what translator and certification process it accepts, then use that process for the version you submit.
Does Khmer text leave my browser when I translate it?
Yes. Pressing Translate sends the source text from your browser to Google’s translation endpoint. If you upload a document for extraction, that file is also sent to InKhmer’s API; image OCR itself runs in the browser before the extracted text is sent to Google. Do not paste material you are not permitted to share with these services, and review the original before relying on the result.
Data and accuracy
Khmer-to-English text is sent by the browser to Google’s translate-pa.googleapis.com/v1/translateHtml endpoint, verified in web-static/app/(marketing)/_custom/_lib/tools/translation/translate-api.client.ts. The page presets km → en in apps/inkhmer.com/data-static/src/data/tools/categories/translation/translate-khmer-to-english/index.ts. The W3C Khmer script resources describe phrase spacing and word-boundary detection. Image OCR runs in-browser with Tesseract.js; document extraction uploads files to InKhmer’s API. The returned text is machine translation and is not certified.
Sources
- Google Translate — Google
- Khmer Script Resources — W3C Internationalization Working Group
- Khmer Honorifics: Re-emergence and Change after the Khmer Rouge — University of Michigan Library
Content last reviewed 21 October 2026