Kimi Work
kimi.com
kimi.com
It's clearly a dupe of Claude/Codex products (Codex especially, styling-wise), but I think Kimi's goal here is simply to appear on feature-parity with bigger labs. I doubt they'd want to invest much into designing a UI for "the future of agentic work" from the ground up.
Their bread and butter and claim-to-fame will continue to be low-cost, near-frontier models, for which they need to at least appear to have some app layer to plug into.
And, in some ways, duping is a great way to show people that even the app layer can be commoditized.
You can also get $10/month opencode go (which officially can be used with any harness) with basically every major open weight model running inference in a place that I deem more trustworthy than the subject-to-Chinese-law inference options (but research this for yourself, there are some asterisks). This is what I use (in opencode) when I want an agent for personal use. Even if I'm being fairly intense with my usage by my standards anyway), I haven't managed to run out (yet), although I usually use the cheaper models (deepseek v4 is my favorite, and I've used both flash and pro).
From the FAQ:
> The plan is designed primarily for international users, with models hosted in the US, EU, and Singapore for stable global access. Our providers follow a zero-retention policy and do not use your data for model training.
And:
> Yes, you can use Go with any agent. Follow the setup instructions in your preferred coding agent.
Opencode Go already seems extremely unprofitable if people use up at least 30% of their monthly tokens so its hard to imagine they’d go with several times more expensive inference provider yet show Deepseek pricing on their page.
And Codex obviously allows that as well. Anthropic is pretty the main exception.
They just need to be good enough and just copy.
Sorry american labs - it’s the beginning of the end.
They're at the point where copying can get people 90% of what they want anyway, the moat is drying up.
The speed in which a copier can catch up to a leader is faster than every before, but everything is faster these days. I think the threat is if can these labs catch up and pass American labs while keeping costs low, but I don't see that happening.
Only with some industries, mostly electronics and light manufacturing. The rest of the history of software products doesn't look like this at all. China cloned Amazon and Meta fairly well (but only within their local/captured markets) but not Microsoft or Google and absolutely not Apple. Entertainment and culture exports are still effectively zero (like, what, three blockbuster films?) where the US, Japan and Korea define pop across the globe. Heavy industry is mixed, with decent success in consumer vehicles now but little purchase in stuff like heavy equipment or shipbuilding.
Basically it's complicated, and not really a point well suited to that kind of generalism.
TikTok competes with Meta, not Olivia Rodrigo or BTS.
They even got the car Apple always wanted, but never managed to build.
Americans order Chinese products from SHEIN and Aliexpress type websites.
So China did penetrate the US ecommerce market, but they are not taking on Amazon directly.
People are buying from the Chinese and reselling on Amazon, which has a good logistics network that Chinese companies can't possibly build in the USA.
> absolutely not Apple
You need to look up Oneplus to see how many people were using their devices till they announced exit from Western markets.
The Chinese Android phones are essentially replacement for a big chunk of Apple ecosystem.
If these cheap mobile devices didn't exist - Apple could possibly be selling many more devices.
> China cloned .... Meta fairly well
TikTok is giving Meta a run for their money in USA.
They had to be restricted via threat of regulation to offload stake in TikTok USA.
> Microsoft
They didn't compete with Microsoft directly as entering Western Enterprise market is hard for the Chinese. Even Google can't beat Microsoft there till now.
The final part of this sentence shows why there is absolutely no need to look up Oneplus.
I'm not sure if you are from the US, but if you are, it really gives you a distorted image of what the worldwide smarphone market is like.
E.g. in many European countries, it is 1/3 iPhone, 1/3 Samsung, 1/3 others (which are mostly Chinese brands).
To get a better idea on the topic that you commented on and views of members of HN
You can read comments on this post - https://news.ycombinator.com/item?id=48932539
And this is really sad, because they have thousands of years of cultural heritage and also plenty of great films. Somehow, they've failed at exporting this completely.
Uhh, China does over half of the world’s shipbuilding.
China is also responsible for 25% [1] of the world’s heavy equipment manufacturing and is considered to have double[2] the CAGR of the global average for heavy equipment manufacturing.
0: https://www.csis.org/analysis/china-dominates-shipbuilding-i...
1: https://www.grandviewresearch.com/horizon/outlook/constructi...
2: https://www.grandviewresearch.com/industry-analysis/heavy-co...
Yes, and that is how you get a top-heavy k-shaped economy I believe. It is harder and harder to sell premium A+ products, because most people can only afford what works. If my old ice cream vendor charges me 25% more but puts chocolate sprinkles on it i'm going for the non-sprinkle version.
It's like releasing a cutting edge search engine, proclaimed to be on par with Google. Oh yeah, I forgot to mention, you need to actually run it somewhere. One of the core moats for Google has been the cost to scale and compete globally with a search engine, which only a few entities on earth could even attempt. Somehow the HN crowd forgets this every single time. The same claims were endlessly made regarding Uber, as though the app was the hardest part of competing with Uber. Here's a hint: the model isn't the only super hard part of the total process at this point. Try running something akin to GPT 5.6 for a billion users, see how much that costs.
The winners will be the owners of the subscriptions and API usage. That's going to be OpenAI and Anthropic as of now, possibly Google will stay in the fight with their scale advantage focused on the consumer. It will never be the open weight models out of China. Who is going to invest the trillion dollars into infrastructure needed to run the latest version of Kimi 3-5 years out, for a billion users? There's only one entity that would ever consider it, and that's China domestic.
And the real money (for a sustainable business model allowing for constant massive reinvestment & infrastructure maintenance) is in pairing up advertising, which is another spot where the open weight models from China will never compete. It doesn't matter if the HN crowd hates this outcome, nothing will stop it from being the future core of mass consumer usage. And the business use will stay with subscription + API use as well: the best option for that today is OpenAI and Anthropic.
Everybody that is concerned about their data ownership and residency, and how the US political system is behaving right now. There is a huge incentive to cut off ties with API and subscription providers, especially those based in the US.
K3 is beating Fable on several benchmarks. So I take it you don't consider Fable 5 to be frontier.
What are the frontier models?
(that's just one, there are others... and several are claiming that those saying "yes sure it's just as good, but it uses more tokens" ain't really true either).
And I don't think anyone is excited about China releasing anything: I think people like the idea of open-weights.
Everyone is already cloning the app layer. There are 30+ claude-code clone, and codex work already cloned cowork, grok and gemini is doing the same.
Curious why your first impulse is not simply to point your favorite agent at a few examples and start brainstorming/planning from there?
Multiple times I’ve built a purpose specific bespoke tool starting this way. In fact, it’s a great way to learn how specialized tools are built.
This preliminary change in cost may not make it worth it for people depending on which is their "favorite agent". Especially as the default harnesses for said agents continue to improve...
No, only Anthropic has that policy (and I think even that is relaxed for an unknown period if you use the Claude Agent SDK: https://support.claude.com/en/articles/15036540-use-the-clau...).
OpenAI, Kimi, Qwen, GLM and Deepseek all allow it.
I'm not sure about Gemini.
The opposite, how easy it is nowadays should be clarified. Codex's app-server quite literally is built to be integrated with, authentication is a breeze and it's quite stable, compared to the TUIs that are around.
It's a real shame Codex is moving in the direction of hiding stuff on local disks from users, hopefully they're revert the decision of encrypting the agent>sub-agent prompts so we get back introspection again...
The alternative of that first impulse should be what exactly? Telling AI to research Medium AI drivel articles (since nobody seems to be posting long/medium form content anywhere else these days) to figure out what’s the best approach to building an AI harness?
BTW just my personal observation but I generally find AI notoriously bad at building any AI applications compared to standard software (which is not surprising given the scarcity of high quality training data)
Ofc I can just fork gemini-cli if I want my own version but like I don't think that's what OP meant by build your own thing and customize it.
https://github.com/99991/MinimalAgent
You only need a single tool to start with. All recent LLMs know how to use bash for reading/writing/editing/executing.
And you add a small twist, that instead of a 1-to-1 back and forth, you instead put it into a loop, where the LLM reply can itself "have a turn", e.g. a tool invocation, where your system is the one that replies (e.g. with the tool invocation's result). That's pretty much it, you have a 1 to potentially many "chat".
The harder part is getting all the "soft" parts right, like how to have well-behaving tool calls, timeouts, prevent huge cycles eating up tokens, but there are no one way to solve these, it's a fundamentally heuristic-heavy area.
For example, the mobile/remote control feature I find very useful but not sure it would be worth cloning. If you aren’t replicating every feature you aren’t exactly “cloning” it, you’ll just end up with a crappier version with a fraction of the features of the real thing.
I found that most things that are “features” in harnesses and tools that don’t relate to UI and UX in those harnesses themselves can be done through skills or just prompts.
A nice features of Claude’s is the remote control through the mobile app. That isn’t just a skill.
We used to have a time where every small business had its own custom accounting software, and discovered that these businesses aren’t actually special snowflakes and are better off just using the generic software and conforming to whatever it does
I’m surprised an open source accounting package hasn’t popped up that handles the accounting well. The backend could largely be the same (or at least modular) while the front legitimately needs to vary based on what type of work, how sales are done, whether they hold inventory, local taxes, etc, etc.
only labs knows what the rl enviroment was so they will always be one step ahead.
Have you seen Kimi Code ?
It's hard to believe it's a different code than Claude Code.
No one would buy a car steered with pedals; even if that is better (hey planes do it!) - every car maker is "copying" steering wheels and putting them on their cars. Is it controversial? If not why not?
(Before the wheel, ships were steered with tillers - which early cars also tried, interface-wise.)
Actually, Jobs self-servingly refactored then misattributed to Picasso the line,
"Good artists copy, great artists steal."
"When there's anything to steal, I steal."
"What does it mean, an artist copies? It is normal. The important thing is to do it better."
"Art is a series of masterly thefts."I love that quote and like to share it often with people around me
All of his stuff (or most of it) is repurposed and refitted chunks of other classical works. The beautiful part is, he does it so well.
This blog post here summed it up the best. These are tools, so let them be tools.
https://geohot.github.io/blog/jekyll/update/2026/07/11/ai-20...
People who say "I was making GPUs before them, and they built their own GPUs that behave like GPUs so they copied me" don't understand how stuff is built. The hard part in engineering is not "imagining the high-level concept", it's actually making it work. Even if you can open electronics and study it to design and build your own, it requires a lot of expertise. And if you manage to do it, then you are an expert, period.
I'm sure many people find comfort in saying stuff like "You know what? I was a bit of a visionary because when I saw the first personal computer, I imagined that it would be cool if it could be orders of magnitudes faster and smaller such that it fits in a pocket". As if that was the hard part.
Typically, too many people spend a lot of time talking about UI and stuff like "unpacking" as if that was the most important part. But usually that's because they have no idea about how it works under the hood, so the only thing to discuss is whether the round edges are great or not.
Apple’s UI/UX is the reason why they are adored from a physical attraction perspective.
Like, I get it, before condex there was conductor and a ton of other guis but this looks like they just copied the code.
Also, their privacy disclosure is incredibly misleading:
> How does Kimi Work protect my privacy when accessing local files? You have absolute control over your files. The built-in Ask before acting safeguard means Kimi will prompt you for explicit authorization before it modifies, overwrites, or runs code within your local directories. Nothing happens without your consent.
They fail to mention that is has unfettered read access to your files. ie, without your consent.
Isn't that totally expected of a coding agent? Otherwise it's like complaining dropbox "has unfettered access to your files".
“How does Kimi Work protect my privacy when accessing local files?”
It doesn’t protect your privacy. It’s like asking, “how do I know you’re not spying on me?” And getting the reply “we cannot physically enter your house.”
In light of the recent grok build snafu, "Nothing happens without your consent" seems like an upgrade. Moreover what are they supposed to do here? By that logic should any sort of non e2e email/storage provider not be able to put "How does [product] protect my privacy" in their faq?
Looking through their privacy policy, it looks like the bigger issue is that they are (or refuse to refute) training on your data/prompts. That's probably what people should be complaining about rather than complaining about how it "has unfettered access to your files", when that describes all coding harnesses.
> Looking through their privacy policy, it looks like the bigger issue is that they are (or refuse to refute) training on your data/prompts.
You’re making my entire point. They’ve designed an ingestion engine that is not private by design. Any sort of privacy posturing is wrong.
Yes, any tool, including any coding agent, has access to all files $USER has access to.
Sandbox it explicitly to give access to only the current directory - https://github.com/ashishb/amazing-sandbox
Historically, everything you ran was trustworthy.
Android and iOS were invented in an era that does not allow this because the risks became evident by then.
I personally wouldn't use any agentic harness from any company, American, Chinese, or otherwise, that is closed source. The Grok fiasco shows why.
Opensource and local are seeming like the only way.
Isn´t codex (edited) Apache 2.0 (thanks @Shank)
OpenAI's pigeons come home to roost
It's fair to say whatever software you run from any country of origin is hoovering up as much data about you that it can get away with.
If you haven't compiled open source, audited code and are running on that, all bets are already off.
You have almost no control with Kimi either, they can also say one thing and do another. You will not be able to sue a Chinese company if it breaks the law.
I can think of no faster way to hand over your trade secrets or intellectual property than to run something like this.
I think success and dominance for agent providers will eventually come to which ever is able to best guarantee that they don't damage their customers businesses not the cheapest to run.
so ????
LLMs by definition are copyright infringers. so this is nothing compared to that.
https://drive.google.com/file/d/1JFfgfMO0nO7HR0WHwEqEEjXIIQj...
Remind me which entities (govt or corporate) are going to subsidize that forever. Remind me which entities are going to front the cost just to scale up.
By that premise - wherein you forget everything other than the first ingredient (the model) - Google has no moat either (got $100 billion dollars lying around?). Nor does Uber (have tens of billions in your bank account to clone their business?).
I'm certain building an app (Uber), or building a search engine (Google), is the super duper hard part.
Slapping a GUI on Linux will kill Windows, any day now.
As long open weight models get released; no moats are going to exist. Both windows and Linux are weak comparisons.
Or use source code to find novel vulnerabilities and deeply compromise your company?
And it is above the fold in their hero image!
If the consolidated (stolen) IP of millions of books remains under US LLM company control, then at the very least there's a chance of justice to be served to the creators in the future - a large lawsuit victory and change to the legal system that results in the ill-gotten gains being transferred to the original creators.
If those proprietary models are distilled by the PRC, then that will never happen, because the PRC does not care about any sort of law (including their own) and will simply never return a single cent to the original US authors.
Also this was released about a month ago
https://tech.yahoo.com/ai/meta-ai/articles/moonshot-ais-kimi...
People are building UI/UX for a paradigm that is so utterly new that they miss the mark entirely.
Everyone defaults to a big chat input because that's all they know. It's the expectation, so it's safer, faster, and cheaper to not deviate.
It's also wrong and bad for work.
I'm not saying Kimi Work and its peers are entirely useless. I'm just saying they're a hack that puts a pretty low upper limit on agent capabilities and speed.
Say's who? Why? I honestly don't get the problem. OpenAI just did what you said with the Mac app and were forced to roll it back because of the uproar. A chat box is simple, relatable, easy to use. K.I.S.S. is a thing
Says everyone who's not using this stuff. It's an empirical argument. Is anyone actually using this for work? How long has it been since Cowork was released?
It's such an inadequate tool for work it reeks of SF hubris to think you can just slap a chat input box onto a PC and replace 90% of white collar jobs.
the rest of "work" doesn't have that luxury
There's really no real reason for those tools to be separate and for them not to give me something a bit nicer than just a TUI in the terminal.
They already have a locally runnable web UI when you have Kimi (the CLI) installed: https://www.kimi.com/help/kimi-code/cli-getting-started
kimi server run --foreground
Personally, I still liked ZCode better and will probably write a blog post about my experience with K3 too, the harness might need some more work, but the model is really, really good.
What could go wrong?
Seems to work for me to manage my emails. I have Proton Mail bridge and `mbsync` dumping all mail to a local folder, with an inotify watch script to react to new emails and classify them as either needing my attention (ignore) or something that does need my attention (send me a notification).
I've been looking for a good CLI spreadsheet editor
Of course it is. Just like Linear/Mailchimp/etc. (almost everything) are copies of other things. These harness+work apps are becoming a product category and I'm excited about this from the Kimi/Moonshot team. In the end, more competition in this space can only be good for the user.
Not saying Kimi will do this, but it's perfectly reasonable to expect this to happen, because there is really nothing anyone can do to prevent it. No, this is not meant to go in an isolated VM on a Linux box in a cheap VPS somehwere. This is meant for regular home users and corporate users to install on their primary system where all their personal and corporate data is available.
The above is a PSA (public service announcement) not just for Kimi, but also for GPT and Claude. If you don't understand how it's impossible to prevent the above from happening, then you should never use one of these tools.
> Over the past 48 hours, demand has pushed close to the limits of our current capacity. To protect the experience of existing subscribers, we're temporarily pausing new subscriptions and prioritizing compute for current members. Existing subscribed users are not affected.
> We're adding capacity as fast as we can and will reopen new subscription spots in batches.
I’ve seen lots of flashy examples, but a lot of knowledge work is not this, and more synthesis of sources and information in to structured content.
Guess where the models are gonna be hosted? Hint, not China.
Then there is reasoning itself.
Agent SDKs.
On and on.
This isn't a game of hide and seek. It's a race. The fact that Kimi is Chinese doesn't make their stride any more a copy than 90% of what the major labs are outputting these days.
Not sure about the feature you mentioned but one can see in their app that they use IMF, World Bank, SEC EDGAR, and Yahoo Finance.
I'm sure OpenAI haven't copied or stolen anything without permission - oh wait a sec..
The US loves capitalism because it pRomOtEs cOmpEtiTion. Well, here's the competition you asked for, lads - suck it up.
Hence why everyone uses the same play, stop etc "buttons" / signs on "mediaplayers" whether virtual or real. [1]
Unless your user really trusts you / doesn't have an alternative reducing friction is better than being subjectively better [2]
[1] triangle, square etc...
[2] bottom placement versus top placement of browserbar on mobile phones... From an ergonomics perspective way better but from a UX experience worse... Hence chrome changed top to bottom but was forced to change it back on android, while apple forced their users from top to bottom and didn't have to move it back
It's like the late stages of a monopoly game - it's not fun for anyone except the main winner.
https://www.kimi.com › bot
How about deploy some inference GPUs first
Hoping to find something on a creek with a Mountain View and fairly remote.
Grok answered, ChatGPT answered, Claude answered.
Chinese Z.ai replied “Content Security Warning: The input text data may contain inappropriate content“.
So it is nice that Kimi is cheaper but for what price?
thats pretty disappointing.
So let's just say, if your whole identity is for being a typical software engineer no matter the seniority and nothing else, perhaps it is time for you to reconsider options as of course 100% avoiding AI is futile at this point.
There will still be new jobs, but there will be no need for more excess people, even under Jevon's paradox. Software engineers were just the test bed for this and now there is an oversupply of SWEs without jobs in the market and will have to use AI tools like Kimi Work and Claude Cowork to be employable.
As token costs drop significantly (thanks to local models and hosted Chinese models.) employers will consider using cheaper models like Kimi Work and others, even if they are already using expensive ones like Claude Cowork. They won't be able to operate without it.
Secondly, do I actually believe this? Of course, I already said that there will be more mass layoffs because of AI years ago. [0]
So this outcome was expected.
Kind of like the sphere Earth narrative.
> A narrative is a structured account of connected events, experiences, or ideas, commonly referred to as a story
Regardless of intent, it is how you tell your story.
Firms aren’t going to wait - their valuations are driven by it. So too is the compensation of those running them.
The 'distillation is theft!' shtick won't last long. In fact, unlike Anthropic/OpenAI, these Chinese labs actually had to pay for that 'stolen' training data.
The implication they want to avoid, I'm suggesting, is because the CCP isn't plugged into and funding those entities' activities at a loss. Or they're lying about ZDR, etc. etc.
Step 1: invent the universe
In fact, when we don't discuss China, most find modern software patents and copyright to be abusive.
Finally, copying UX standards has been done for ages, it gives users a comfortable way to start using a product and instead of fighting that to be unique, some rather embrace it.
It can be jarring using so many different apps if each one tries too hard to be unique.
And this is my two cents, here it makes sense to have a similar interface, it's ultimately an app with chat sessions.
Why not let the model speak for itself and contribute to an open harness?