const translator = await Translator.create({ sourceLanguage: 'en', targetLanguage: 'fr', });
await translator.translate('Where is the next bus stop, please?');
That said, the "simple way around the paid API" problem is something Google has to deal with anyway, because there are a bunch of ways to use Google Translate without paying for it (e.g., the translate.google.com web UI, or the non-JavaScript-exposed "Translate to [language]" feature built into Chrome), and any action that can be taken by a human can in principle also be taken by a script. The only thing to do about it is use secret-sauce server-side abuse detection to block as much of this activity as they can; they can't get all of it but they can get enough to push enough people onto the paid API that the economics pencil out.
Also, it says to have 22 GB free, but below (under "Note"...), it says the model takes "around a couple of GB".
(Answered my own question): Doesn't look like it requires the user's permission. Upon first use, the model will start downloading. The user has to wait for the download to finish before the API will work. That could take hours for 22gb.
I presume this can't work on mobile?
https://developer.mozilla.org/en-US/docs/Web/API/Translator_...
It runs very slow. Test case that run translation for text in 3k chars multiple times, takes about 30 seconds for google translator, but more than 10 minutes for `nllb-200-distilled-600M`.
Text sample: https://github.com/translate-tools/core/pull/112/files#diff-...
My tests runs on nodejs, it looks in browser it have no chance for real world use
Basically, the 'translate this' button you see on Twitter or Instagram next to comments in foreign languages. This API would make it trivial for all developers to add that to their web apps.
- it's experimental
- the "specification" is nowhere near a standards track: https://webmachinelearning.github.io/translation-api/
Of course it's already shipped in Chrome, and now Chrome pretends that its own Chrome-only API is somehow standard. Expect people on HN to blame other browsers for not shipping this.
The W3C is not a prescriptive standardization body. It doesn't have any regulatory power giving it any teeth to go after vendors acting in bad faith. So the W3C process is descriptive and encourages a period of competitive divergence in implementations. It is only after the early adopters have hammered on the features and figured out which parts they like best that a Web API can then start to get standardized.
Let me quote the site for you
--- start quote ---
This specification was published by the Web Machine Learning Community Group. It is not a W3C Standard nor is it on the W3C Standards Track.
--- end quote ---
> So the W3C process is descriptive and encourages a period of competitive divergence in implementations.
That is exactly opposite of how the w3c standardization process works
> It is only after the early adopters have hammered on the features and figured out which parts they like best that a Web API can then start to get standardized.
Yes, and until then this work is not supposed to be enabled by default
Being "standards track" means the spec is out of draft and has been proposed. It does not mean "we intend to standardize this". It means, "we've put in all of the work to standardize this and are waiting on final acceptance".
I don't know what you mean by "isn't supposed to be enabled by default". There is no mention of when browser vendors may or may not ship features in the standardization process.
The literal "Draft Community Group Report" (and not a working draft) is a literal link to w3c standardization process: https://www.w3.org/standards/types/#CG-DRAFT
Since the words "not on the W3C Standards Track" from the document didn't persuade you, you could go to the actual w3c process and answer a few simple questions:
- is "Draft Community Group Report" a document on a standards track?
- what does it take to get on the standards track?
- what does it take to "put in all of the work to standardize this and wait on final acceptance", and how many steps there are between "Draft Community Group Report" and this stage?
> I don't know what you mean by "isn't supposed to be enabled by default".
For a person who is so confidently talking about the w3c standards process, I'm surprised you don't.
w3c doesn't explicitly state this. Except for the final few stages, all steps in the process contain the following: "Software MAY implement these specifications at their own risk but implementation feedback is encouraged."
However.
Since this is browsers we're talking about, it means that whatever browsers ship enabled by default will remain in the wild forever because people will immediately start depending on that implementation.
Additionally, a standard cannot become a standard until there are at least two independent implementations of a proposed feature. This is to eliminate the possibility to ship purely internal APIs, or depend on a single library/implementor.
So the way to do it, especially for APIs that are nowhere close to being "waiting for final acceptance" is: ship behind a flag, iron out issues and differences, perhaps change the API (and changes to API happen all the time), then ship.
Of course, Chrome shits all over this process and just ships whatever it wants to ship.
No "however". The W3C is not actually a standards body. They don't get to tell people what to do. The W3C even knows this, even though we all colloquially call these things "standards", they're actually just "Recommendations".
The FCC is a standards body. NIST is a standards body. You go against them, you get fined. Messing with weights and measures is one of two crimes defined in the US Constitution, they both come with the death penalty, and the other one is treason.
That's not the W3C. You go against the W3C, worst case scenario, Apple says, "nah, we ain't gonna do that" and then Web devs don't adopt your feature because they can't run it on one of the biggest platforms: Safari on iOS, the only browser allowed to run on iOS.
What you've just pointed out is the W3C explicitly saying it doesn't mind if browser vendors implement features early. They "MAY" do it. And then they remind everyone the risk is on the vendor if the eventual sta... excuse me, "recommendation", diverges from what does eventually get standardized.
Oh look. You've stopped claiming that this spec is on the standards track, that the final step is just final acceptance, or that there's an actual standardisation process.
You've now switched to saying that this is not an actual standardization process and to pretending that I said something I never did: that w3c enforces standards.
Imagine if you actually knew anything about what you were talking about and argued in good faith.
Adieu.
If this API will be implemented in next few years, there will be browsers who hold up the world in progress.
Linguist have enough many users, so we could expose this API for client side code, and users with browsers where Translation API is not implemented yet, could install Linguist and sites who uses Translation API would works fine. Translation API calls would proxy by Linguist and preferred translator module would be used.
Any thoughts about it?
Safari/webkit is positive (though no official stance yet):
https://github.com/WebKit/standards-positions/issues/339#iss...
> The spec assumes a certain form of translation backend, exposing information about model availability, download progress, quotas, and usage prediction. We'd like to minimize the information exposure so that the implementation can be more flexible.
reads to me as Chrome once again trying to export itself verbatim as a "standard" and Mozilla pointing out that that's not really applicable to others.
Also the WebKit post seems to raise somewhat similar arguments but on the basis of fingerprinting/privacy problems.
Why not just use the lang= attribute as it was intended, then let the user select text to translate as they wish?
Instead, a JS API gives more flexibility and control.
Besides, I think the "lang" attribute is supposed to signal what the language of the text inside that element is, not what it could/should be. So even if going with attributes would be the way forward, a new one would need to be created.
I get that this is one more brick in the wall that teams like LadyBird will have to maintain, but as a web developer I do think more Web API features is generally a good thing, as it makes it easier for smaller shops to implement richer functionality.
I assume every browser will do the same as on-device models start becoming more useful.
Last time I looked I wasn't able to find any easy to run models that supported more than a handful of languages.
You can use llm for this fairly easily:
uv tool install llm
# Set up your model however you like. For instance:
llm install llm-ollama
ollama pull mistral-small3.2
llm --model mistral-small3.2 --system "Translate to English, no other output" --save english
alias english="llm --template english"
english "Bonjour"
english "Hola"
english "Γειά σου"
english "你好"
cat some_file.txt | english
Plus, mistral-small3.2 has too many parameters. Not all devices can run it fast. That probably isn't the exact translation model being used by Chrome.
https://github.com/facebookresearch/fairseq/tree/nllb/
If running locally is too difficult, you can use llm to access hosted models too.
Not the easiest, but easy enough (requires building).
I used these two projects to build an on-device translator for Android.
https://ollama.com/library/gemma3
> support for over 140 languages
You could also look into Argos Translate, or just use the same models as Firefox through kotki [4].
[0] https://huggingface.co/facebook/nllb-200-distilled-600M [1] https://huggingface.co/facebook/m2m100_418M [2] https://huggingface.co/google/madlad400-3b-mt [3] https://huggingface.co/models?other=base_model:quantized:goo... [4] https://github.com/kroketio/kotki
Worst case scenario a user-script/extension could monkey patch it out, but probably clients will let you disable it.
Rather, it's an API developers can use to add inline translation to web apps.
For example, under a comment in your app, you can (a) detect the language, and (b) if it's different from the current users/browsers language, offer to translate it with a small link (c) if the user clicks the link, the content is translated to their language.
Bacause it was, is, and will be Chrome-only for the forseeable future: https://news.ycombinator.com/item?id=44375326
right?