frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Cursor Composer 2 is just Kimi K2.5 with RL

https://twitter.com/fynnso/status/2034706304875602030
100•mirzap•1h ago

Comments

Sammi•1h ago
As a paying customer, it just doesn't feel good that they are trying to pass off someone else's model as their own.

I mean I guess this is what businesses do all the time. There's a term for it even, it's called white-labeling.

But is this all that Cursor have? They pass of VS Code as their own, they pass off Kimi as their own... What do Cursor even do? What do I need them for?

jstanley•1h ago
As a paying customer, I don't care where the model comes from, I only care how good it is.
Sammi•1h ago
Sure, and also at what price point.

But can I rely on Cursor to be able to keep delivering, when they aren't the one's doing the work themselves?

khuey•1h ago
White-labeling may be slightly dishonest to the consumer but the manufacturer and distributor are honest with each other. That doesn't appear to be the case here (Kimi's license requires publicly acknowledging Kimi is used for anyone operating at Cursor's scale).
justindotdev•1h ago
im pretty sure this is in violation of moonshot's ToS. this is going to be fun to watch unfold
gillesjacobs•1h ago
They probably licensed it. Still a bit deceptive not to mention it on the model card/blog post, but companies whitelabel all the time without mentioning.

It goes against the ML community ethos to obscure it, but is common branding practice.

deaux•1h ago
No they didn't [0][1]. With this leak they're probably negotiating as we speak, which could be why they've deleted the posts.

[0] https://chainthink.cn/zh-CN/news/113784276696010804

[1] https://pbs.twimg.com/media/HD2Ky9jW4AAAe0Y?format=jpg&name=...

charcircuit•1h ago
Kimi K2.5 was released under a modified MIT license (100M+ MAU or $20M+ MRR has to prominently display Kimi K2.5). It will be fine.
antirez•57m ago
Basically this is true for most startups in the world BUT Cursor, so here you are kinda inverting the logic of the matter. Cursor is at a size that, if they wanted to use K2.5, they could clearly state that it was K2.5 or get a license to avoid saying it.
NitpickLawyer•53m ago
IF we assume that the modified MIT clause is enforceable. And if we assume Cursor Inc. is running the modification. It could very well be the case that Cursor Research LTD is doing the modifications and re-licensing it to Cursor Inc. That would make any clause in the modified MIT moot.
NitpickLawyer•1h ago
There is no ToS at play here. There's only the license[1], which is MIT modified like so:

> Our only modification part is that, if the Software (or any derivative works thereof) is used for any of your commercial products or services that have more than 100 million monthly active users, or more than 20 million US dollars (or equivalent in other currencies) in monthly revenue, you shall prominently display "Kimi K2.5" on the user interface of such product or service.

[1] - https://huggingface.co/moonshotai/Kimi-K2.5/blob/main/LICENS...

zozbot234•51m ago
Yes, this is pretty clear-cut. There's even a great alternative, namely GLM-5, that does not have such a clause (and other alternatives besides) so it feels a bit problematic that they would use Kimi 2.5 and then disregard that advertisement clause.
NitpickLawyer•48m ago
I've replied down the thread, but there are ways to go around that clause entirely, even if it would be enforceable. The obvious way is to have another company do the modification.
zozbot234•42m ago
The worthwhile question AIUI is whether AI weights are even protected by human copyright. Note that firms whose "core" value is their proprietary AI weights don't even need this (at least AIUI) since they always can fall back on "they are clearly protected against misappropriation, like a trade secret". It becomes more interesting wrt. openly available AI models.
kgeist•11m ago
At the same time, Moonshot violated Anthropic's ToS by training on their models' outputs :) [0]. And Anthropic violated copyright law by training on copyrighted material. It's violations all the way down.

https://news.ycombinator.com/item?id=47126614

rvz•1h ago
Honestly I don't think this leak is any good for Cursor. Not only this appears as a violation to Moonshot's ToS, this may also be in fact enough evidence for Anthropic to ban Cursor from using their models, just like they are doing to OpenCode.

Why? As I said before, Anthropic mentions Moonshot AI (Maker of the Kimi models) as one of the AI labs that were part of this alleged "distillation attack" [0] campaign and will use that reason to cut off Cursor, Just like they did to OpenAI, xAI and OpenCode.

Let's see if the market thinks Composor 2 is really that good without the Claude models helping Cursor. (If Anthropic cuts them off).

[0] https://www.anthropic.com/news/detecting-and-preventing-dist...

charcircuit•1h ago
Kimi K2.5 is an open source model. It is intended for people to make derivative models.
gillesjacobs•1h ago
Cursor is mostly an IDE / coding-agent harness company. So it probably makes sense for them not to train their own base model, but instead license something like Kimi and fine-tune it for their own harness and workflows.

Their moat looks pretty thin. A VSCode fork with an open-source LLM fork on top. In the fast-moving coding-agent market, it’s not obvious they keep their massive valuation forever.

jstummbillig•1h ago
There is a plausible scenario in which software engineering requires a very finite amount of intelligence, in which sota models will be used mainly for other things and where for coding the harness will become increasingly more important than the model.
NitpickLawyer•55m ago
> Their moat looks pretty thin.

Their value is in the data they've collected and are collecting. Usage, acceptance rate, and all the connected signals. Plus having a large userbase where they can A / B test any finetune they create.

_puk•45m ago
There were conversations in the team yesterday about how Cursor's cloud agents are still ahead of Claude from a UX perspective.

Obviously we're running both, using the right tool for the job.

There is stickiness there from being early. That will be hard to replicate.

CharlieDigital•39m ago
That's every harness including VC Code Copilot.

People home about Teams sucking, but its market share is several times that of Slack because of distribution.

I guarantee that Microsoft has even more data.

agluszak•1h ago
A hyped startup providing zero added value, burning investor money only to repackage somebody else's work? That's new... /s
deaux•1h ago
Looks like two Moonshot employees confirmed that it's not licensed before Moonshot made the decision to get out of the debate and delete their posts [0][1].

[0] https://chainthink.cn/zh-CN/news/113784276696010804 - may have originally been https://x.com/apples_jimmy/status/2034920082602864990

[1] https://pbs.twimg.com/media/HD2Ky9jW4AAAe0Y?format=jpg&name=...

lmc•57m ago
This is on their website...

"Is Kimi K2.5 open source?"

"Yes, Kimi K2.5 is an open source AI model. Developers and researchers can explore its architecture, build new solutions, and experiment openly. Model weights and code are publicly available on Hugging Face and the official GitHub repository."

https://www.kimi.com/ai-models/kimi-k2-5

saidmukhamad•46m ago
4th paragraph in license block

Our only modification part is that, if the Software (or any derivative works thereof) is used for any of your commercial products or services that have more than 100 million monthly active users, or more than 20 million US dollars (or equivalent in other currencies) in monthly revenue, you shall prominently display "Kimi K2.5" on the user interface of such product or service.

lmc•34m ago
That's not an open source license, then.
bakugo•31m ago
It wouldn't be regardless, because the model is open weights, not open source. It's just a license.
lmc•25m ago
Which contradicts what they say on their website.
ffsm8•21m ago
My first reaction was "well, who knows how much revenue they're actually doing"

But at least the rumor mill has them significantly above that line:

> Revenue: As of March 2026, reports suggest Cursor has surpassed $2 billion in annualized revenue (ARR).

rfoo•42m ago
TBH they really shouldn't have posted such a tweet in the first place, just sit back and watch their license enforced by the Internet.

I had the question "how do you even enforce this weird license term" back then, I guess I know the answer now.

lossolo•1h ago
Their first model was also based on an open source Chinese base model. They never fully trained their own model.
todteera•58m ago
From a users perspective, do we really care what model we're using under the hood? Or how well the software is solving our problems?

Seems like cursor is trying to build a "thicker wrapper" beyond the harness. Either to protect against Anthropic shutting them off or increase margins.

mohsen1•56m ago
Cursor Composer 1 was Qwen and this is Kimi. IDE is based on VSCode. The entire company is build on packaging open source and reselling it.

Ollama is also doing this.

There is so much money to be made repackaging open source these days.

So funny to see Twitter go wild saying "a 50 person team just beat Anthropic" blah blah.

NitpickLawyer•49m ago
> packaging open source and reselling it.

It's a bit more than that. They have plenty of data to inform any finetunes they make. I don't know how much of a moat it will turn out to be in practice, but it's something. There's a reason every big provider made their own coding harness.

pbowyer•39m ago
Can anyone enlighten me how having a coding harness when for most customers you say "we won't train on your code" helps you do RL? What's the data that they rely on? Is it the prompts and their responses?
rubymamis•38m ago
I guess they rely on many people not toggling privacy-mode on?
victorbjorklund•21m ago
I doubt the majority does that. I bet the majority is using the defaults.
rubymamis•38m ago
Do you know what Qwen model Composer 1.5 used?
olejorgenb•49m ago
To be fair, is "with RL", "just"?

They should have disclosed it though. If they didn't it's a bad look for sure.

mono442•48m ago
This whole ai stuff feels like a big bubble especially with the oil price soon at $200 and guaranteed recession.
koakuma-chan•46m ago
Cursor can't compete with Claude Code's subsidized pricing, so they are trying to gaslight people that their cheap model is good enough.
vachina•34m ago
A question. I’m due for a yearly Cursor subscription renewal, how does the credit limit look like?

Currently I’ve not hit any of the limits despite using it quite rigorously, I wonder if this will change with a renewal?

DeathArrow•30m ago
I whish it was GLM 5.0.
prodigycorp•29m ago
There are many reasons to make fun of Cursor. However, one of the things get right is their autocomplete model.

Are there any open models that come close? Why doesnt OAI or Anthropic dedicate some resources to blowing Cursor's model out of the water? Cursor's completion model is a sticking point for a lot of users.

olejorgenb•4m ago
The model is great. The UX is horrible...
heliumtera•16m ago
For all the muh productivity guys that like to claim they can turn invisible when no one is looking, an produce 600k lock over 6 weeks, well...cursor is useless now. We know kimi K2.5 won't make you 100 trillion times faster.

Cursor is killed for this market.

cbg0•4m ago
Scores higher than Opus 4.6 on their in-house benchmark? Sounds legit.

Floating wetlands boost water quality, slash greenhouse emissions

https://www.science.org/content/article/floating-wetlands-boost-water-quality-slash-greenhouse-em...
1•Brajeshwar•42s ago•0 comments

Is Discord Down? Users Report Errors

https://comuniq.xyz/post?t=871
1•01-_-•1m ago•0 comments

Some critical issues with the SWE-bench-Pro environments

https://github.com/SWE-agent/mini-swe-agent/issues/787
2•snoopyswe•2m ago•0 comments

Is my job safe from AI? (version 4)

https://99helpers.com/tools/is-my-job-safe-from-ai
1•nickk81•3m ago•1 comments

Evilwaf – MitM Firewall bypass proxy and WAF vulnerability scanner

https://github.com/matrixleons/evilwaf
1•matrixleons•4m ago•1 comments

Oil stored at sea is running down fast [ ] 1/3 of what's left is Iranian

https://www.bloomberg.com/news/articles/2026-03-20/oil-market-s-seaborne-buffer-runs-down-fast-as...
1•alecco•4m ago•1 comments

A Trillion Transactions [video]

https://www.youtube.com/watch?v=y2_BqkKTbD8
2•sgt•5m ago•1 comments

Middle East Conflict Is Rewiring Global Supply Chains

https://www.eetimes.com/middle-east-conflict-is-rewiring-global-supply-chains/
2•giuliomagnifico•9m ago•0 comments

Show HN: triad.js – A toolset for creating web apps

https://codeberg.org/joexo/triad.js
1•joexo•11m ago•0 comments

Six Roads to World War III

https://thekochicompass.substack.com/p/six-roads-to-world-war-iii-b69
1•shreshthdugar•14m ago•0 comments

Garmin data archive – local, offline, Ollama-ready

https://github.com/Wewoc/Garmin_Local_Archive
1•Wewoc•14m ago•1 comments

100+ Kernel Bugs in 30 Days

https://substack.com/home/post/p-188916866
3•someplaceguy•15m ago•0 comments

My Own Personal Claws

https://www.bstjohn.net/blog/claws/
1•danab3791•15m ago•0 comments

Paul Graham, Founder, Y Combinator [video]

https://www.youtube.com/watch?v=2Q2uh1BlqKA
1•tim333•19m ago•1 comments

Probiotic cream that ramps up heat production could prevent frostbite

https://www.newscientist.com/article/2520006-probiotic-cream-that-ramps-up-heat-production-could-...
2•Brajeshwar•20m ago•0 comments

Claudy Day

https://pages.oasis.security/rs/106-PZV-596/images/claudyday-vulnerability.pdf?version=0
1•jeisc•23m ago•1 comments

The Reason Windows Hate Is Exploding: It's the End of Personal Computing [video]

https://www.youtube.com/watch?v=e7a89ZYcTo8
2•oldnetguy•24m ago•0 comments

Introducing DoorDash Tasks

https://about.doordash.com/en-us/news/introducing-doordash-tasks
1•pramodbiligiri•25m ago•0 comments

Molly Holzschlag's archived blog remake

https://mollyholzschlag.com/
1•vietyork•25m ago•0 comments

Show HN: I built a super simple email reminder inspired by the RemindMe! bot

1•santah•26m ago•0 comments

Whiplash: When people stop taking GLP-1s

https://nautil.us/whiplash-heart-attack-and-stroke-risk-jumps-when-people-stop-taking-glp-1s-1279029
1•I_Nidhi•27m ago•0 comments

A static site generator and website transferring under 20kB

https://theconsensus.dev/blog/2026/03/19/a-static-site-generator.html
2•ibobev•29m ago•0 comments

The Rise of Fake Casio Scientific Calculators

https://hackaday.com/2025/12/29/the-rise-of-fake-casio-scientific-calculators/
1•bookofjoe•29m ago•0 comments

AI (2014)

https://blog.samaltman.com/ai
23•bjornroberg•29m ago•6 comments

Kanban TODO: a textfile based Kanban board in a single HTML file

https://www.youtube.com/watch?v=yyBZceJG-Ls
1•nialse•30m ago•0 comments

A visual way to explore startups that raised recently

https://iraised.io
1•damiannn•33m ago•0 comments

Missile strikes close to reporter during Israeli attacks in Lebanon

https://www.aljazeera.com/video/newsfeed/2026/3/19/missile-strikes-close-to-reporter-during-israe...
4•0x54MUR41•34m ago•0 comments

Surveillance in Africa: Mapping Chinese AI Surveillance Across 11 Countries

https://www.ids.ac.uk/publications/smart-city-surveillance-in-africa-mapping-chinese-ai-surveilla...
1•giuliomagnifico•35m ago•0 comments

AI's impact on mathematics is analogous to the car's impact on cities

https://mathstodon.xyz/@tao/116252708577614828
2•birdculture•36m ago•0 comments

Entso-E final report on Iberian 2025 blackout

https://www.entsoe.eu/publications/blackout/28-april-2025-iberian-blackout/
4•Rygian•36m ago•0 comments