Chrome extensions and tools for translators, and what each one saves

How browser extensions speed up translation work

If you or your team translate in the browser, your extensions decide two things: how fast a draft moves, and how clean it is when it reaches a client document or a live page. A fast stack saves time on every job. A careful one keeps a client’s trust, the harder of the two to earn.

The stack most people rely on has also moved on. Chrome changed how extensions are built, several favourites were retired with the change, and AI assistants now do much of what three or four translation tools used to do. A list bookmarked in 2022 is due a refresh.

Below is the toolkit we keep installed in 2026: what each extension saves, what we dropped and how the tools fit into a checked translation workflow.

What changed since the last version of this article

Two shifts reshaped the old toolkit, and both change where your time goes. First, Chrome retired Manifest V2 between 2024 and 2026. Chrome disabled the remaining Manifest V2 extensions for all users on 24 July 2025, and Google removed those still listed from the Chrome Web Store on 31 August 2026. Extensions had to migrate to Manifest V3 to keep working: several translation extensions rebuilt themselves around service workers, and others left the Chrome Web Store.

In May 2024, as Google announced the start of the Manifest V2 phase-out, over 85% of actively maintained extensions in the Chrome Web Store were running Manifest V3.

Source: Google, "Manifest V2 phase-out begins", The Keyword, 30 May 2024

The second shift is bigger. Large language models moved into the browser through dedicated extensions and sidebar panels. For everyday lookup, quick rephrasing and post-editing of machine output, one LLM extension now does what three or four translation extensions used to do.

The global language services and technology industry generated $49.68 billion in 2023, a 4.5% drop from $52.01 billion in 2022, driven by enterprise adoption of neural machine translation and large language models.

Source: CSA Research, 2024 Market Sizing Update, 6 September 2024

Nimdzi, which sizes the market on its own method, estimates the language industry at $72.6 billion in 2025 and projects $73.4 billion for 2026, growth of under 1%.

Source: Nimdzi, The 2026 Nimdzi 100, April 2026

The two firms count different things, so each total is best read on its own terms. Both show why a faster, checked workflow matters to a translator working under price pressure.

Translation engines we still keep on the toolbar

A good engine gets you to a usable first draft fast. The choice is which one to trust for which job.

DeepL for Chrome

Best raw quality for European languages, in our daily experience. We reach for DeepL first when we need a French or German draft that should sound natural. Select text on any page, press a shortcut, and read the translation in a popover right where you are.

The free tier covers most quick lookups; Pro unlocks document mode and glossaries in the main app. DeepL also owns Linguee, so a paid plan ties context examples and engine output together in one workflow.

ImTranslator

We keep ImTranslator installed for one job: comparing translations from Google, Microsoft Bing and other engines side by side in the same popup. A three-way comparison quickly exposes a phrase that needs a human decision.

Mate Translate

The on-page bubble translator we default to for casual reading. Select a word, get the definition and translation, and save it to a phrasebook synced across devices. Pro adds Netflix subtitle translation, useful when checking how a streaming service rendered a colloquial phrase.

We use Mate as a word and phrase helper and send whole documents through the checked workflow below.

Google Translate

Every client recognises it, which is its main strength. Clients send us websites and expect us to read them quickly, and the official Google Translate extension is the cleanest way to get the gist of any page in seconds.

AI assistants that quietly replaced half the old toolkit

The errors that matter most sit in tone, register and ambiguity more than in single words, and those need context, which an LLM extension can take. For complex sentences, idiomatic content, or anything legal or technical, a general-purpose LLM extension now beats most dedicated translation tools.

Claude in Chrome and ChatGPT

We can paste a paragraph plus three lines of surrounding context and a one-line instruction such as "translate to formal French for a Belgian law firm audience, keep the second-person plural form". An LLM extension takes that instruction as written; a standalone engine works from the text alone.

We use Claude in Chrome, available on paid Claude plans (Anthropic still lists the Chrome side panel as beta), for client work where tone and register matter, and ChatGPT for quick rewrites and brainstorming. We proof every output from both by hand. See our deeper view on how to use AI and machine translation tools for the workflow we run them through.

DeepL Write

Sits halfway between a translator and a grammar tool, rephrasing your target-language draft for tone, register and natural flow. We run our own French and Spanish drafts through it, then accept or reject each suggestion sentence by sentence. It catches the stiffness that comes from translating word for word in your head.

Reference helpers we use every day

Most translation choices come down to picking between two plausible meanings. These lookup tools settle that quickly, and they are how you check an AI draft before you rely on it.

Linguee

A bilingual concordancer with real example sentences mined from EU documents, patents and parallel corpora. When we want to double-check a DeepL translation, Linguee shows how the term was used in twenty real documents. It is now part of DeepL but still works as a standalone reference.

Reverso Context

Same idea as Linguee, with different source material: stronger for informal speech, film subtitles and conversational register. Between the two, you have working examples for almost any phrase.

LanguageTool

A multilingual grammar and style checker for French, Spanish, German, Dutch and more. Grammarly is English-only at the depth we need, so for any non-English work LanguageTool is the better pick.

Two more tools for translators who also handle web content

If your translation work touches websites, two more extensions let you quote with a full view of the site and confirm the source is sound before you localize it.

Wappalyzer tells you what CMS, plugins and translation setup a site runs before you open the source code. When a prospect asks about translating their WordPress site, we want to know in two seconds whether it runs WPML, Polylang, TranslatePress or a custom system. Our website localization service page explains what each one means in practice.

Detailed SEO Extension gives a fast on-page audit of headings, metas, hreflang, canonical tags and schema markup. Before we localize a site into three more languages, we want to see whether the source-language SEO is competent. Our round-up of Chrome extensions for SEO covers the wider audit toolkit.

What we removed from the previous version of this list

Every extension you keep is another popup, shortcut and permission, so we keep the list lean. These five made way.

Readlang Web Reader. Useful for language learners, whose day it fits best.

TransOver. Its engine quality has dropped behind DeepL's, and the pop-up trigger conflicts with several modern sites.

Rememberry. Flashcard work sits best in Anki.

Grammarly. Still excellent for English-only writers; for multilingual professionals, LanguageTool covers more languages in depth and replaces it here.

Lingvanex. Still works, but DeepL plus an LLM extension covers the same ground with better output.

Compare the extensions side by side

DeepL Translator lists more than 100 languages; DeepL Write's text improvement covers nine: Chinese, English, French, German, Italian, Japanese, Korean, Portuguese and Spanish (checked 26 September 2026).

Source: DeepL API documentation, Languages supported

The biggest change in over two decades of multilingual work is not the quality of MT or even the rise of LLMs. It is that the browser became the workbench. Everything we used to do across Trados, a paper dictionary, and three monitors now happens inside one Chrome window with seven extensions. The job is the same. The tools are unrecognisable.

Mike Bastin, multilingual SEO and translation consultant

How we sequence these in a working translation day

The tools pay off in the right order: machines for speed first, a human for accuracy last. A French legal document translated into English might pass through the five checks below.

1. Create a first draft in DeepL.
2. Flag legal terminology and check each term of art against Linguee.
3. Use Claude in Chrome to test ambiguous sentences against the agreed brief.
4. Run the English through LanguageTool to catch grammatical drift.
5. Read every sentence aloud and verify it against the source.

Where the translator's browser stack is heading

Sidebar LLMs are likely to absorb more word-level lookup, while bilingual corpora such as Linguee and Reverso Context remain useful for checking model output against real examples.

CAT tool companions are also moving more of the workflow into the browser. For a wider view, see our piece on how AI is transforming translation and localization.

The browser stack is the surface. Underneath is a workflow built to check MT and LLM output faster than the per-word price drops. If you handle multilingual content across markets and want a second opinion on the production setup behind it, see how our agency approaches expert translation services and post-AI editing, or get in touch and we can walk through your current stack in 20 minutes.

Frequently asked questions