Getting exhausted with the constant technical problems on here

I love Lingq. I have used it for years. It is in my view the only app that actually works, when you put in the work. The problem is only the never-ending technical problems.

After being down (as it is so often) yesterday, today I tried the extension on Netflix to import subtitles, and it seems to just import snippets from the video, I believe (I cannot relate the text to the actual video at all).

Lingq is methodically sound, but the actual technology around the method is a mess. I have yet to encounter any other subscription based website that goes down or loses functionality as often as this website. The fact that it is a subscription based, paid service should create the expectation that it is a professionally maintained site.

Just a little bit of venting after years of daily, loyal and enthusiastic usage.


I agree. It is a real shame that there are so many technical problems with a service that’s so valuable at its core.


Unfortunately, this is a sad reality. And it is also an expensive service. It would be definitely awesome if they could find a way, and invest some money, to create a stable product, with a stage/testing environment, where customers could have a reliable experience.
I don’t use, or I limit myself by using various functionalities to reduce that frustration.


Similar impressions here.

I’m a software architect by background and do software product strategy for a living now.

Although a daily user of LingQ for now several years, I do have limited outside-in perspective. That said, I’m confident I can read between the lines quite a bit.

I’d surmise that LingQ used to have a technical problem and now they have a business problem. The list of technical things to-do has gotten out of hand, there’s just simply too many things to do. Thus, at this point, it’s not a matter of numerous technical decisions but one of a few business decisions.

In this context, it probably can’t go unsaid that Steve is 78. He needs annuity income, not a money pit.

As far as I know, there isn’t anything else quite like LingQ. I personally immediately became a fan. I recall that 25 years ago, for Japanese, I coded an assisted reader myself to be able to click on any word in browser to quick get the dictionary definition. That said though, if I look ahead to the future, even the near term, the use of generative AI at the core of the architecture of an open comprehensible input content platform is what’s necessary.

Not only is LingQ now technologically fragile, it’s also ready to be overcome by what’s next, generationally. And what’s next isn’t just technical and generative AI, isn’t also a more robust framework of enabled incentives and money flows for second language content creators.

Right now, there’s a wedge for some of the best and the brightest, in LingQ and in the LingQ user community along with others savvy in generative AI, language acquisition, and two-sided platforms, to envision a new ground-up platform. Likely some of those individuals come across this post.

LingQ’s successor needs launched.

I don’t say any of this to be unsupportive. I’m LingQ’s biggest fan but with an open transparency toward prospects and potential, at least as I see them.

LingQ, you have the right of first refusal to be the world’s best and dominant language acquisition platform. The global 21st century needs what you’ve envisioned.

Simply though, the world will move ahead with or without you at the helm, or on the boat, or not.


Amen to that. I love LingQ too. The concept is brilliant.

However, the implementation is slipshod and buggy – the worst software I’ve ever paid for. Most free software is better.

How about just getting the keyboard shortcuts to work reliably?


If Language Reactor were to make certain changes and implement some of their already planned upgrades then LingQ would be royally screwed.


I believe this is something Language Reactor definitely has right; the immersive experience is in the content–in the video, in the article, in the browser, in the context of, really, the other brand, immersive and ambient.

I’m not saying LingQ’s plug-in-based “import” metaphor is wrong. I’m saying, it’s not enough and the future center.


Today I am just particularly frustrated with Lingq. I have resubscribed today to Netflix for the sole purpose of importing shows to Lingq. I would otherwise not be interested in Netflix. So there goes a tenner (to quote Kate Bush). Afterwards, I find out that the Netflix import doesn’t even work any more. So not exactly money well spent.

So I return to a book on Lingq that I have imported some time ago and get really invested in it. After reading the first half, I notice, that several parts of the import got lost, while others got duplicated.

Then I realize that this happened to a lot of older imports and that those courses are no longer complete. Lingq lost the data it seems. Unfortunately I no longer have the original file.

So eager was I to finish the book, that I was ready to buy it again. Another tenner spent on behalf of Lingq today. But alas, it is no longer available for purchase in my target language. So now I can simply not finish the book.

Furthermore, Lingqs are no longer correctly counted and have not been for months now. Whenever I create a Lingq it gets added to the number of Lingqs in that lesson, but when I mark a Lingq as known it doesn’t get subtracted.

Therefore, I do not actually know what number of Lingqs any given lesson has. Considering there was also a time when the words read count was highly inflated, my words read count is probably completely out of whack by now.

In summary, Lingq just doesn’t do anymore, what it advertises to do. It doesn’t import from Netflix, it doesn’t keep or display relevant statistics, it doesn’t even save your data.

The reading function itself still works, fortunately, provided that Lingq hasn’t lost the text itself.

I would presume that these constant hassles are a big reason, why Lingq, despite being a better method than any other language app or website, never became a part of the internet language learning mainstream like Duolingo did.

You just cannot recommend Lingq to the genral public who would just be confused and eventually annoyed by it.


Anybody know anything about Memrise?

1 Like

Have you tried the @roosterburton extensions? I use them constantly for youtube and Netflix and find them excellent.

I realise it’s not ideal to have to use 3rd party addons but they do add value.

@Caldazar @roosterburton’s addons also have an editor with a find/replace feature which works very well.


I’m with you! I love LingQ in principle, but there are things that are just infuriating. The library, for instance, is almost unusable. I can’t meaningfully view things, and it is ridiculous that I can’t arrange what I do and don’t want to see, that I can’t make it only show me natively-hosted stuff, that it keeps giving me stuff above my requested level, and the list goes on. I hate the library view.

It also sucks that we can’t fix lessons anymore. Maybe too many people screwed it up, but it used to be that if I found a typo in a lesson that was clearly a typo, I could fix it. But back then they also didn’t underline the approximate line of text the audio was reading, which might be a reason not to allow just anyone to correct things.

Reviewing flashcards is also difficult without a lot of tinkering. I get that Steve hates flashcards, but I can’t imagine why he wouldn’t ensure that those who want the functionality can’t just easily review unknown random words a a particular importance level or whatever.

I love LingQ, but if I can find something that does the same thing better, I’ll jump ship. It should be better. I wish it were better.


I have tried the Rooster Burton extensions.

Would be awesome if his ideas were fused with more investment! I’m a fan.


I have been a Memrise user for quite a few years with a paid subscription. For me, the main advantages were to have the ability to use user-generated content/courses for less popular languages (dead or alive, real or artificial) (e.g. Georgian, Sumerian) and create my own courses, which I could share with others.

Memrise has changed course and is now concentrated on what is described in your shared article. They deleted the user forum at the end of last year. As far as I have understood, many »older« users have moved on to Anki as a learning tool. Somebody even cloned the Memrise community courses to a different website as they are supposed to be deleted from Memrise itself (see this Reddit post). At the moment, the community courses still exist on the Memrise website.

I have not used the new app as they offer the usual languages, which is certainly understandable as that is what the majority wants and probably needs.


I very much like the discussion I see in this topic. I don’t agree with everything here, but the general direction of the discussion is very knowledgeable and accurate. Here are a few comments.

tech problems: I also see a large number of bugs all over the place, but I have applied the Stephen Krashen’s acquisition-learning hypothesis to the learning of how the Lingq website works, and I now know how to avoid getting frustrated.

comparing lingq bugs to bugs of other software: For me, Lingq is somewhere in the middle, for how well it manages bugs, compared to other software (software of all kinds: games, financial stuff, etc). But simply because they are in the middle with regard to quality doesn’t mean they have a sound business model. I worry about the Lingq business model.

can or cannot fix lessons: I have been fixing a handful of lessons Jan-Feb this year, with no difference in behavior from what I did in mid 2023. It works pretty well.

How soon AI will turn everything upside down: I use clozemaster, and they allow a monthly quota of “explain this sentence” AI queries. I tried it for Arabic. The AI seems about as smart as a classmate who gets better grades than me, and with a super fast dictionary available to them. And it gives an entirely different kind of information from what a dictionary lookup gives, which is very nice. I can basically learn grammar (!) from scratch, just by clicking on the “ask AI” button.
clozemaster AI is not as smart as an average human teacher, yet. About one out of 20 times, the AI, without indicating that it is joking, says something quite wrong. Balancing positives and negatives, and making some wild guesses based on paying some attention to Lex Fridman, Sam Altman, etc, this gives me the feeling that AI will turn the online no-human-teacher language learning market upside down within less than a year.


Re:fixing lessons, I think certain lessons are “locked” (akin to Wikipedia articles that get protected) – those may or may not be all of the lessons provided by LingQ itself, I’m not sure on that.

As for AI, I still find it very unreliable for language explanations, particularly languages and/or grammar/vocabulary points that are niche enough for a Google search not to turn up the answer. (Which makes sense, given that the AI would need to get the answer from somewhere itself. But the problem is that instead of saying it doesn’t know, it hallucinates an answer, sometimes even inventing intricate grammar rules that do not exist.)

1 Like

I haven’t found any lessons that I can fix since 5.0 went live. Granted, I don’t find typos all that often, but I haven’t been able to fix any of them.

And thanks to the library view being terrible, it’s hard to find native LingQ content I’m interested in anyway, which doesn’t help.


I’ve been trying the free version of Memrise & have been kind of frustrated with the interface.


1 Like

Give up. Let it go. Next time LingQ frustrates you, sing to yourself:

What is off?
LingQ don’ hurt me,
don’ hurt me,
no more

And move your head like this.

Don’t let 'em put you down anyway.


As far as I am concerned, I mainly use LingQ as an ebook reader and I think it works pretty well, although it would be nice if further improvements were made.

However, I find the communication between the team and the users to be virtually nonexistent. We have no idea what is going on behind the scenes, what bugs are fixed or not, what features we can expect to see…

The forum is an absolute mess. The team should be able to apply labels to topics to indicate whether this is being worked on or has been fixed already, just like a roadmap does.

They have a significant number of developers working for them, but truth be told, I am not sure their skills are being used effectively.

I also find that the team is quite reluctant to bring new changes that are being requested by users of the platform. They will only improve what they see fit and the rest falls by the wayside.


This is a thread about complaints, and there’s nothing wrong with that. But I just want to say that I forgive a lot of the small problems, because I see integrity around what they’ve provided. And that integrity is not such a common thing in a commercial company.

As a point of comparison, consider DuoLingo. I know that the features aren’t apples-to-apples with what LingQ people want, but I’m just using a product that many people here will be familiar with for comparison.

DuoLingo has far fewer bugs than LingQ. It gamifies very well and delivers those dopamine hits that keep you coming back to the language “gym”. The UI is gorgeous. It’s got the new AI stuff. And believe it… they nailed the business model. In short, they addressed many complaints in this thread, to the point where people here might ask, “should LingQ be more like DuoLingo?” in how it manages its product.

But when I’m using DuoLingo, I feel suspicious of its slickness. I’m a subject in a psychology experiment. I’m a rube that’s wandered into a language learning casino. Friendly and useful things like their community forums get disappeared to make room for a new AI-enabled paid tier. They revise the content structure so often that it’s impossible to predict when you’ll ever finish. And in my more cynical moments, I believe they’d rather I never finished, and kept coming back to their site forever.

With LingQ, I never have that manipulated feeling that so much software creates. I like that the Lessons page has no fancy algorithm that predicts what I’m going to like next. I like that a lot of the content is community-driven. I’m happy about the API not being protected, so I can hack little tools together based on LingQ data. I like running into hackers like Rooster with his nifty plugins. I like that having a critical conversation like this one on the official forum is fine.

But I wonder if the economics of software create a dilemma between integrity/under-funded and manipulative/well-funded. I’ve made software for years and years, at my own companies and other peoples’. This conflict has come up repeatedly in my career.

Imagine how Steve, Mark, and the team would find the money to improve LingQ dramatically…

Some venture capitalist sees potential in LingQ and wants to take it to the next level. They bring in millions of new funding, hire up some extra devs, install an ex-DuoLingo product manager to set the roadmap for fixes and features. Bugs you’re complaining about get fixed. Everything is A/B-tested with feature flags and all the decisions that make the site sticky and keep people paying are put in place. Some features get cut. Some features get added. The VC funds are attached to investors that want 100x return on their investment, not 2x. And LingQ turns into another language learning casino–very profitable, but kinda icky. “Candy Crush” your second language.

I hope for something in the middle for LingQ. A nice medium-sized company, comfortably profitable, a little less bugs, spirit intact.