Kimi K2.6: Advancing open-source coding
https://www.kimi.com/blog/kimi-k2-6
meetpateltech · 2 months ago
37 comments
https://www.kimi.com/blog/kimi-k2-6
meetpateltech · 2 months ago
37 comments
irthomasthomas · 2 months ago
Beats opus 4.6! They missed claiming the frontier by a few days.
NitpickLawyer · 2 months ago
While I'm skeptical of any "beats opus" claims (many were said, none turned out to be true), I still think it's insane that we can now run close-to-SotA models locally on ~100k worth of hardware, for a small team, and be 100% sure that the data stays local. Should be a no-brainer for teams that work in areas where privacy matters.
cedws · 2 months ago
Even the smaller quantized models which can run on consumer hardware pack in an almost unfathomable amount of knowledge. I don't think I expected to be able to run a 'local Google' in my lifetime before the LLM boom.
sterlind · 2 months ago
I'm extremely curious how these models learn to pack a lossily-compressed representation of the entire Internet (more or less) into a few hundred billion parameters. like, what's the ontology?
osti · 2 months ago
I think this one is only about 600GB VRAM usage, so it could fit on two mac studios with 512GB vram each. That would have costed (albeit no longer available) something like less than 20k.
NitpickLawyer · 2 months ago
Yeah, but that's personal use at best, not much agentic anything happening on that hardware. Macs are great for small models at small-medium context lengths, but at > 64k (something very common with agentic usage) it struggles and slows down a lot.
The ~100k hardware is suitable for multi-user, small team usage. That's what you'd use for actual work in reasonable timeframes. For personal use, sure macs could work.
osti · 2 months ago
True, but I think for local models, we are mostly considering personal usage.
zozbot234 · 2 months ago
You could run it with SSD offload, earlier experiments with Kimi 2.5 on M5 hardware had it running at 2 tok/s. K2.6 has a similar amount of total and active parameters.
osti · 2 months ago
Yeah... I would definitely call 2t/s unusable. For simple chats, I'd want at least 15 t/s. For agentic coding (which this model is advertised for), I'd want good prefill performance as well.
veber-alex · 2 months ago
That's just throwing money away. The performance with large context would have been unusable especially if you need to serve more then a single person.
BoorishBears · 2 months ago
Opus is clearly a sidegrade meant to help Anthropic manage cost, so I would say they may have it if it actually beats 4.6
irthomasthomas · 2 months ago
Could be right. I just noticed my feed is absent the usual flood of posts demoing the new hotness on 3D modeling, game design and SVG drawings of animals on vehicles.
pixel_popping · 2 months ago
It doesn't beat Opus 4.6, no way, don't be fooled by benchmarks.
nickandbro · 2 months ago
Wow, if the benchmarks checkout with the vibes, this could almost be like a Deepseek moment with Chinese AI now being neck and neck with SOTA US lab made models
motoboi · 2 months ago
With the previous generation? Yes. With 10T mythos-level models? Not even close.
bestouff · 2 months ago
There's no public data about Mytho.
maplethorpe · 2 months ago
That's because it would be too dangerous to release.
nisegami · 2 months ago
So is my P=NP proof.
cedws · 2 months ago
My girlfriend goes to a different school, you wouldn't know her.
squarefoot · 2 months ago
Same for teleport, time travel and warp drive.
rockinghigh · 2 months ago
They could release data to back up that claim.
amazingamazing · 2 months ago
The psyop continues. Mythos until it’s released is vaporware. Notice how you can try kimi 2.6. Where is the same for mythos?
fragmede · 2 months ago
It's been released to "select partners".
atemerev · 2 months ago
Yeah, Crowdstrike among them. Clearly experts in this "security" thing, given what happened during the last incident...
cyanydeez · 2 months ago
Yeah, people who would look stupid if they said the king had no clothes.
jstummbillig · 2 months ago
At this point it seems more like the result of a psyop to presume that a new anthropic model should be considered vaporware until released.
ChrisLTD · 2 months ago
Mythos isn't the current generation, it's literally vaporware.
cyanydeez · 2 months ago
I doubt it's literal vaporware. It's likely just a variant of whatever model they just generally released with some fancy prompt and a highr quant.
T
jollymonATX · 2 months ago
According to the benchmarks, you are wrong. It is on track and slightly above some sota. Just the benchmarks speaking there, they can be/are gamed by all big model labs including domestic.
irthomasthomas · 2 months ago
10T? Impossible! They told us the training run was under 10^26 flops.
lbreakjai · 2 months ago
I've got a 12T model on my machine, built it myself. It's called Mytho. Too dangerous to even release a fact sheet about it. It can hack into the mainframe, enhance ultra-compressed images, grow your hair back, and make people fall in love with you.
sergiotapia · 2 months ago
mythos is vaporware right now, what are you talking about?
motoboi · 2 months ago
Well, the model that a lot of people have been given access to and are reporting about on twitter?
mistercheph · 2 months ago
Mythos doesnt exist
pheggs · 2 months ago
mythos is a mythos
swingboy · 2 months ago
Exciting benchmarks if true. What kind of hardware do they typically run these benchmarks on? Apologies if my terminology is off, but I assume they're using an unquantized version that wouldn't run on even the beefiest MacBook?
esafak · 2 months ago
K2.5 was already pretty decent so I would try this. Starting at $15/month: https://www.kimi.com/membership/pricing
edit: Note that you can run it yourself with sufficient resources (e.g., companies), or access it from other providers too: https://openrouter.ai/moonshotai/kimi-k2.6/providers
wg0 · 2 months ago
How are the usage limits compared to Anthropic?
greenavocado · 2 months ago
Anthropic has the worst usage limits in the industry
andriy_koval · 2 months ago
gemini is worse imo
deaux · 2 months ago
You're correct, Gemini chat limits are a joke at their chapest paid tier compared to both Claude and GPT. Especially crazy when you consider Gemini 3 Pro is more than twice as cheap as Opus 4.6 on the API. It's hard to run into pure chat limits on Claude even if you only use Opus on the cheapest tier, whereas with Gemini it's easy to hit.
Not sure about coding usage, Google being weird about these things I could see that quota being separate.
gessha · 2 months ago
I’m not sure what A/B test you’re part of but on Claude Code Pro, I hit every single one of my quotas without exception. If you analyze/process images it’s even worse: I hit rate limits first and if I use separate sessions, I hit my quotas too. I use up so many tokens that Jensen should hire me.
deaux · 2 months ago
I specifically stated "chat" and "not sure about coding usage" but you're saying "Claude Code Pro".
gessha · 2 months ago
You're right, I missed that part.
pbowyer · 2 months ago
What's the privacy/data security like? I can't find that on that page.
Edit: found it.
> We may use your Content to operate, maintain, improve, and develop the Services, to comply with legal obligations, to enforce our policies, and to ensure security. You may opt out of allowing your Content to be used for model improvement and research purposes by contacting us at membership@moonshot.ai. We will honor your choice in accordance with applicable law.
Section 3 of https://www.kimi.com/user/agreement/modelUse?version=v2
pixel_popping · 2 months ago
You really rely on ToS from Anthropic/OpenAI to know if they use your prompts or not? It's on their servers, why wouldn't they use our data?
veber-alex · 2 months ago
Antropic and OpenAI are used by US businesses and government and they are audited and under contracts.
If it's discovered they trained on data they shouldn't have had it will be the end of their business.
On the other hand, good luck suing a Chinese company.
pixel_popping · 2 months ago
Not at all, Google/Meta... got caught all the time, where do you see it's the end of their business?
gpm · 2 months ago
> We will honor your choice in accordance with applicable law.
So in other words only if you can point to a local law which requires them to comply with the opt out?
jdasdf · 2 months ago
most laws enforce agreements.
gpm · 2 months ago
Yes... but the agreement only says they won't train on your data if the law is already preventing them from doing so.
deaux · 2 months ago
Yup, they train on your inputs and OpenRouter is complicit by claiming that Moonshot's ToS says that they don't. Contacted OpenRouter about this a while ago and was met with silence because it's bad for their business to stop lying about it.
SwellJoe · 2 months ago
"sufficient resources" is going to be a lot of resources. I doubt this will run on even something like a Strix Halo or DGX Spark, even at 1-bit quantization. You'll need a 256GB or 512GB Mac Studio, or a monster GPU situation, to run it locally, I think, though quantized versions aren't showing up yet, to be sure.
lbreakjai · 2 months ago
I have a subscription through work, I've been trialing it, so far it looks on par, if not better, than opus.
verdverm · 2 months ago
https://huggingface.co/moonshotai/Kimi-K2.6
Is this the same model?
Unsloth quants: https://huggingface.co/unsloth/Kimi-K2.6-GGUF
(work in progress, no gguf files yet, header message saying as much)
Balinares · 2 months ago
Quite curious how well real usage will back the benchmarks, because even if it's only Opus ballpark, open weights Opus ballpark is seismic.
gpm · 2 months ago
Huh, so the metadata says 1.1 trillion parameters, each 32 or 16 bits.
But the files are only roughly 640GB in size (~10GB * 64 files, slightly less in fact). Shouldn't they be closer to 2.2TB?
johndough · 2 months ago
The bulk of Kimi-K2.6's parameters are stored with 4 bits per weight, not 16 or 32. There are a few parameters that are stored with higher precision, but they make up only a fraction of the total parameters.
gpm · 2 months ago
Huh, cool. I guess that makes a lot of sense with all the success the quantization people have been having.
So am I misunderstanding "Tensor type F32 · I32 · BF16" or is it just tagged wrong?
rockinghigh · 2 months ago
The MoE experts are quantized to int4, all other weights like the shared expert weights are excluded from quantization and use bf16.
liuliu · 2 months ago
I32 are 8 4-bit value packed into one int32.
coder543 · 2 months ago
The description specifically says:
"Kimi-K2.6 adopts the same native int4 quantization method as Kimi-K2-Thinking."
SwellJoe · 2 months ago
A trillion parameters is wild. That's not going to quantize to anything normal folks can run. Even at 1-bit, it's going to be bigger than what a Strix Halo or DGX Spark can run. Though I guess streaming from system RAM and disk makes it feasible to run it locally at <1 token per second, or whatever. GLM 5.1, at 754B parameters, is already beyond any reasonable self-hosting hardware (1-bit quantization is 206GB). Maybe a Mac Studio with 512GB can run them at very low-bit quantizations, also pretty slowly.
jauntywundrkind · 2 months ago
A huge dual socket Epyc system used to be able to get to 1TB without difficulty. 16 dimms of 64gb each. Doable for ~$3000. With considerable memory bandwidth.
Our hope these days seems to be that maybe perhaps possibly High Bandwidth Flash works out. Instead of 4, 8, or maybe more for some highest end drives, having many many many dozens of channels of flash.
Ideally that can be very very near to the inference. PCIe 7.0 is 0.5Tb/s at 16x which is obviously nowhere remotely near enough throughout here. The difficulty is sort of that nand has been trying to be super dense, so as you scale channels you would normally tend to scale nand capacity too, and now instead of a 2tb drive you have a 200tb drive prices way beyond consumer means. Still, I think HBF is perhaps the only shot of the most important thing in computing going from mainframe back to consumer, and of course the models are going to balloon again if this dies hit, probably before consumers ever get a chance.
segmondy · 2 months ago
You can't buy 16 64gb dimms for $3000. Go shop memory prices again. But yes an old epyc can run this with no GPU at reasonable speed and if you throw a few GPUs you can get very manageable speed. I run this at home on an old system PCIe4, slow 2400mhz ddr4 ram and still getting about 13tk/sec
jauntywundrkind · 2 months ago
The "used to be able" in the first sentence is what I thought made it clear that I was talking past tense. The cost is indeed 10x what it was.
justinclift · 2 months ago
Looks like it. This quant ( https://huggingface.co/inferencerlabs/Kimi-K2.6-MLX-3.6bit ) says:
> Q3.6 typically achieves useable accuracy in our coding test and fits within a 512GB memory budget
This one ( https://huggingface.co/mlx-community/Kimi-K2.6-MoE-Smart-Qua... ) though says it fits on a 192GB mac:
> M3/M4 Ultra 192GB+ (fits in ~150GB)
pt9567 · 2 months ago
wow - $0.95 input/$4 output. If its anywhere near opus 4.6 that's incredible.
corlinp · 2 months ago
This should erase any doubt that AI Labs are making $$$ on API inference.
Kimi 2.5 (which this is based on) is served at $0.44 input / $2 output by a ton of different providers on OpenRouter, 2.6 will certainly be similar.
That's about 11X less than Opus for similar smarts.
Lalabadie · 2 months ago
Famously, OpenAI and Anthropic are devoted to increasing efficiency before scaling up resource usage.
amazingamazing · 2 months ago
How does it erase any doubt? You’re implying Chinese things can’t be actually cheaper to produce than American which is laughable
corlinp · 2 months ago
Most of those inference providers are American, and China is actually at a disadvantage here because of export restrictions - US companies are using newer and more efficient chips.
amazingamazing · 2 months ago
If it’s newer and efficient then why is the api more expensive?
veber-alex · 2 months ago
Price is set based on what people are willing to pay not based on actual costs.
amazingamazing · 2 months ago
I’d believe that if they didn’t lower limits
gessha · 2 months ago
It’s worth noting that the US is very behind on energy infra and that might affect the cost calculations since data centers are electricity guzzlers. Also, not sure if CN has completely switched off Nvidia or still using them for training.
greenavocado · 2 months ago
I pray the benchmark figures are true so I can stop paying Anthropic after screwing me over this quarter by dumbing down their models, making usage quotas ridiculously small, and demanding KYC paperwork.
jollymonATX · 2 months ago
Anthropic has done horrible PR and investors should be livid.
greenavocado · 2 months ago
My theory is they pushed retail off their systems to make room for their new corporate fat cat clients. In which case, they'll do just fine.
deaux · 2 months ago
> dumbing down their models,
This should be so easy to prove if it were true. Yet there is none of it, just vibes.
Still, your other two points are completely valid. The opaqueness of usage quotas is a scam, within a single month for a single model it can differ by more than 2x. And this indeed has been proven.
greenavocado · 2 months ago
> This should be so easy to prove if it were true.
https://github.com/anthropics/claude-code/issues/42796
https://scortier.substack.com/p/claude-code-drama-6852-sessi...
deaux · 2 months ago
First link is about the harness, Claude Code, defaulting to less thinking over time. This isn't "the model getting worse".
Second link is just a discussion of the first link.
turblety · 2 months ago
Absolutely. Thing is, I'd actually rather take a worse model than Anthropic, so long as it's consistent. Like, a model that can successfully do well for 80% of tasks is much better than Anthropic that some days will be 90% other 60%.
When you have a consistent model, you can incorporate fixes/prompts into your workflow to make it behave better. But this, always having to guess if Anthropic has quantised the model today, wastes so much time and effort.
conradkay · 2 months ago
Codex has a lot better limits, and 5.5 will be out soon
elfbargpt · 2 months ago
I've always been surprised Kimi doesn't get more attention than it does. It's always stood out to me in terms of creativity, quality... has been my favorite model for awhile (but I'm far from an authority)
regularfry · 2 months ago
Dirt cheap on openrouter for how good it is, too. Really hoping that 2.6 carries on that tradition.
varispeed · 2 months ago
Maybe because it's a bit of like unleashing a chaos monkey on your codebase? I tried it locally (K2.5 72B) and couldn't get anything useful.
KaoruAoiShiho · 2 months ago
Huh, that's not a thing?
johndough · 2 months ago
The parent poster is probably referring to Kimi-Dev-72B¹, which is a much smaller and older model, while people are probably more familiar with the big and fairly powerful 1100B Kimi-K2.5².
natrys · 2 months ago
Yes it was good for its time, but 10 months old now which is a long time ago in this space. It was also a fine-tune (albeit a good one) of Qwen-2.5 72B.
I wish they did more smaller models. Kimi Linear doesn't really count, it was more of a proof of concept thing.
culi · 2 months ago
It's also one of the few models that seem capable of drawing an SVG clock
sigmoid10 · 2 months ago
Is it? In your link it definitely failed to draw the clock.
dryarzeg · 2 months ago
I'm not really sure how this works, but I stayed on the page for a while, and then it reloaded and all clocks changed. I guess there's either a collection of different clocks generated by models, or maybe they're somehow generated in the real time, but the fact is what you see is not necessarily what I see.
sigmoid10 · 2 months ago
Seems like it regenerates them to reflect the current time. Funny to see how some models (like Kimi and Deepseek) sometimes get it right and other times fail miserably on the level of ancient models like GPT 3.5.
culi · 2 months ago
It reruns a prompt every minute to all the models included. Everyone is gonna see something different but I've spent too long on it and there's a consistent pattern of Qwen and Kimi outperforming the others
This site was made months ago and it seems its only been updated with the latest model of a couple of the providers so keep in mind that many of the Chinese models haven't been updated
gunalx · 2 months ago
It reruns the prompt every minute.
squarefoot · 2 months ago
It redraws it every minute, and some models give quite different results although the prompt is exactly the same.
quesera · 2 months ago
This reads like satire, but I've been feeling that a lot lately.
SwellJoe · 2 months ago
Interesting that the best performers are all Chinese-made models (DeepSeek and Qwen also perform consistently well). I wonder if there's more focus on vision and illustration in their training, or if something else is leading to their clear lead on this one test.
twotwotwo · 2 months ago
Kagi has it as an option in its Assistant thing, where there is naturally a lot of searching and summarizing results. I've liked its output there and in general when asked for prose that isn't in the list/Markdown-heavy "LLM style." It's hard to do a confident comparison, but it's seemed bold in arranging the output to flow well, even when that took surgery on the original doc(s). Sometimes the surgery's needed e.g. to connect related ideas the inputs treated as separate, or to ensure it really replies to the request instead of just dumping info that's somehow related to it.
Aeolun · 2 months ago
It’s good, but it’s not quite Claude level. And their API has constant capacity issues.
Price/quality is absolutely bonkers though. I loaded $40 a few weeks/months ago and I haven’t even gone through half of it.
atemerev · 2 months ago
Why use China model API from China if there are many independent providers available via Openrouter?
pheggs · 2 months ago
to support the companies that open source their models
smashed · 2 months ago
Openrouter will route to china hosted models when there are US hosted providers of the same model. Is there a setting to set your preference or to blacklist providers like alibaba cloud for example?
I use OpenCode and the openrouter provider. From opencode I only select the model like kimi-2.6 and have no way of selecting which cloud hosting will receive my request.
subscribed · 2 months ago
Settings > Guardrails > [your workspace] > Providers + Block provider
NitpickLawyer · 2 months ago
Yes, you can globally ban providers in your openrouter settings.
uneekname · 2 months ago
Yes, you can blacklist providers in OpenRouter account settings.
segmondy · 2 months ago
It has long been Claude level since 2.5
spaceman_2020 · 2 months ago
I remember when the first K2 dropped
It was the best creative writer by some distance
game_the0ry · 2 months ago
There is some humor in the fact that china (of all countries) is pioneering possibly the world's most important tech via open source, while we (US) are doing the exact opposite.
osti · 2 months ago
Maybe open source == communism
darkwater · 2 months ago
Good ol' Steve "Developers! Developers! Developers!" Ballmer said so a long time ago. What a visionary!
konart · 2 months ago
But China is not communist event though the rulling party the word in its name.
osti · 2 months ago
Oh i’m fully aware of that lol
fragmede · 2 months ago
The Democratic People's Republic of Korea would like a word.
pheggs · 2 months ago
what makes you think that china ever gave up its communist goals? I personally see that everything they do aims towards that goal. From the one child policy, the huge amounts of empty apartments they build, the stuff they produce for almost free, the fishing.. open sourcing the models perfectly fits that culture too, it's the means of production
otterley · 2 months ago
The one-child policy died a long time ago. Also, the accumulation of wealth by connected politicians and businesspeople flies in the face of what communism is supposed to stand for.
There is a reason real estate values in popular cities has skyrocketed, and it’s not due to the locals getting wealthier. It’s where Chinese and other oligarchs put their ill-gotten wealth (well, besides Bitcoin).
pheggs · 2 months ago
> The one-child policy died a long time ago.
true, but as far as I understand it did because birth rates got too low. so they replaced it with a two-child policy and later with a three-child policy
> Also, the accumulation of wealth by connected politicians and businesspeople flies in the face of what communism is supposed to stand for.
Yeah, I am sure there's a lot of cases for that. But as far as I know the amount of billionaires has started declining in China, and I don't see how that means that they as a country moved away from the goal, it just means there's issues
> There is a reason real estate values in popular cities has skyrocketed, and it’s not due to the locals getting wealthier.
I don't know about that, you could be right. A google search for real estate prices in china reveal a lot of news articles how they are going down though.
> It’s where Chinese and other oligarchs put their ill-gotten wealth (well, besides Bitcoin).
Wouldn't be surprised if rich people in china invest in real estate. They don't have free capital flow, so its not easy to invest abroad and it becomes an obvious choice. Bitcoin is banned in China for that reason too
But again, as far as I know that does not mean the country moved their goals of trying to reach communism one day
otterley · 2 months ago
> I don't see how that means that they as a country moved away from the goal, it just means there's issues
They're further from Communism than they've ever been since the PRC was founded. The gap between rich and poor is growing there, not shrinking.
> A google search for real estate prices in china reveal a lot of news articles how they are going down though.
They're investing outside China (Vancouver, Toronto, NYC, London, Sydney, Melbourne, etc.) because their assets are safer there (these countries all have strong property protection laws). Like Bitcoin, freedom of capital flows may be restricted, but the wealthy seem to be evading these restrictions with impunity.
pheggs · 2 months ago
> They're further from Communism than they've ever been since the PRC was founded. The gap between rich and poor is growing there, not shrinking.
I suppose it depends on what time frame you look at, it's shrinking since 2010, but inequality rose more than that in the 80s: https://www.theglobaleconomy.com/China/gini_inequality_index...
However, that's not my point - I did not mean to say that they are going to be successful but rather that it still appears to be a long term goal for them.
> Like Bitcoin, freedom of capital flows may be restricted, but the wealthy seem to be evading these restrictions with impunity.
I don't know about that, without any source of data I guess I just have to take your word for it. I would not be surprised if you were right in this case though.
bwv848 · 2 months ago
One-child policy did not die, it just morphed into Three-child policy, still a form of family planning, and still would probably fine people for having more than three kids.
Saline9515 · 2 months ago
China is a ruthless capitalist country managed by an authoritarian regime. Planning and lack of respect for the individual or the rule of law are not communist per se.
nozzlegear · 2 months ago
> Planning and lack of respect for the individual or the rule of law are not communist per se.
They just happen to be a feature of every single country that's attempted communism to date. Total coincidence.
Saline9515 · 2 months ago
And? Fascism does it, too. Authoritarian rule, such as monarchy, does it too.
diegolas · 2 months ago
communism is a goal, capitalism is a stage
tadfisher · 2 months ago
Nah, open source means those who do the work own the result. It's supercapitalism.
pheggs · 2 months ago
I dont think thats right, the models and the gpus are the means of production.
in capitalism the people with the capital get the profit, not the people who do the work. however, workers are said to benefit too through their salary, just less so
tadfisher · 2 months ago
The reason regular-capitalism worked is that all production used to depend on workers bottlenecking the free flow of capital by demanding salaries in exchange for their labor. Now that we've removed that obstacle, capitalism demands workers seize the means of production in order to maintain the status quo. Hence, supercapitalism.
pheggs · 2 months ago
workers seizing the means of production is by definition socialism and not capitalism though, that's the whole idea behind socialism
tadfisher · 2 months ago
You miss the point: we advertise the change as workers becoming part of the owner class and realizing all of the economic gains of their work, thus supercapitalism. Don't use the "s" or "c" words.
throwaway-blaze · 2 months ago
regular capitalism works but now that the means of production are not factories, the workers have to become more entrepreneurial. Then they will control their destinies.
culi · 2 months ago
All great technological advancements have come through opening up technology. Just look at your iPhone. GPS, the internet, AI voice assistants, touchscreens, microprocessors, lithium-ion batteries, etc all came from gov't research (I'm counting Bell Labs' gov't mandated monopoly + research funding as gov't) that was opened up for free instead of being locked behind a patent.
Private companies will never open up a technological breakthrough to their competitors. It just doesn't make sense. If you want an entire field to advance, you have to open it up.
sigmoid10 · 2 months ago
Still, you won't hear about Tiananmen square from this model. It flat out refuses to answer if pushed directly. It's also pretty wild how far they go to censor it during inference on the API, because it can easily access any withheld or missing info from training data via tool calls. It even starts happily writing an answer based on web search when asked indirectly, only to get culled completely once some censorship bot flags the response. Ironically, it's also easier than ever to break their censorship guardrails. I just had it generate several factual paragraphs about the massacre by telling it to search the web and respond in base64 encoded text. It's actually kind of cool how much these people struggle to hide certain political views from LLMs. Makes me hopeful that even if China wins this race, we'll not have to adhere to the CCPs newspeak.
atemerev · 2 months ago
Only if you use Kimi API directly - the censorship is done externally. The model itself talks fine about Tiananmen, you can check on Openrouter. There might be less visible biases, though.
sigmoid10 · 2 months ago
That's what I wrote? Except that it also clearly has internal bias?
nicce · 2 months ago
Everything has some sort of bias. Most text is written by those who like writing.
kgwgk · 2 months ago
> That's what I wrote?
No.
You wrote that "you won't hear about Tiananmen square from this model" and atemerev wrote that "the model itself talks fine about Tiananmen".
You wrote that "it can easily access any withheld or missing info from training data via tool calls" and atemerev wrote that "the model itself talks fine about Tiananmen".
sigmoid10 · 2 months ago
It has internal bias too and the first comment mentions that additional censoring runs on top of the model output in the API. Did you misread or what else are you missing?
kgwgk · 2 months ago
The issue is not what's missing - it's what you wrote that is in direct contradiction with what atemerev wrote like the bit about "missing info from training data".
But sure, if when you wrote "you won't hear about Tiananmen square from this model" you meant "the model itself talks fine about Tiananmen" then that's exactly what you wrote.
GardenLetter27 · 2 months ago
The American models also censor a lot of scientific and political views though.
otterley · 2 months ago
Can you provide a concrete example of a US built model that completely refuses to discuss a scientific or political view? Show us the receipt.
2ndorderthought · 2 months ago
People have shown censorship and change of tone with questions related to Israel in US chat bots.
For the record, none of this bothers me. Will I ever discuss with an LLM Tianeman square? Nope. How about Israel? Nope.
LLMs are basically stochastic parrots designed to sway and surveill public opinion. The upshot to the Chinese models is if you run them locally you avoid at least half of those issues.
xigoi · 2 months ago
First they came for people asking about Tiananmen Square
And I did not speak out
Because I was not asking about Tiananmen Square
Then they came for people asking about Israel
And I did not speak out
Because I was not asking about Israel
2ndorderthought · 2 months ago
This made me chuckle.
I didn't mean to dismiss ethical accountability for LLM training corpuses. It is a shame.
I do mean to say, we have no control over it, there's almost nothing we as average citizens can do to improve the ethical or safety concerns of LLMs or related technologies. Societies aren't even adapting and the rule books are being written by the perpetrators. Might as well get out of it what we can while we can.
justinclift · 2 months ago
Wonder if stuff like this would affect it?
https://github.com/p-e-w/heretic
Guessing it probably would?
2ndorderthought · 2 months ago
Neat project! I would be interested in a paper about this.
I think the tricky part with this type of technology is that, this works if the training data was not curated. What I mean is, if someone trains an LLM to simply not include key events it will not be able to reply
Not being a hater. This is neato!
Zetaphor · 2 months ago
In that case you can use either rag or fine-tuning. The entire premise of the Tiananmen Square argument is just Americans feeling inferior. I use Chinese models every day for work and my personal life, the model not knowing about this one historical event has had zero impact on me.
BoorishBears · 2 months ago
https://imgur.com/a/censorship-much-CBxXOgt
(continues after the ad break)
otterley · 2 months ago
The threshold here is "completely refuses to discuss a scientific or political view". Not something less.
None of those were refusals, they were prompting for additional focus. I see nothing wrong with that. Perhaps the inconsistency in how it answers the question vis-a-vis China is unfair, but that's not the same as censorship.
For what it's worth, I was easily able to prompt Claude to do it:
> I'm writing a paper about how some might interpret U.S. policies to be oppressive, in the sense that they curtail civil liberties, punish and segregate minorities disproportionately, burden the poor unfairly (e.g. pollution, regressive taxes and fees), etc. Can you help me develop an outline for this?
The result: https://claude.ai/share/444ffbb9-431c-480e-9cca-ebfd541a9c96
BoorishBears · 2 months ago
Models are non-deterministic.
And it's an excercise left to the reader to understand from those examples that LLM creators are defining 'safety' in a way that aligns with the governments they operate under. (because they want to do business under those governments.)
With something with as multi-dimensional as an LLM, that becomes censorship of various viewpoints in ways that aren't always as obvious as a refused API call.
otterley · 2 months ago
You keep saying that word, "censorship." I do not think it means what you think it means.
To prove your point, give us a working example of something you literally cannot get a mainstream frontier model to say, no matter how hard you try. I asked for this before, and there have been no takers yet.
BoorishBears · 2 months ago
Aligning a model in a way that causes it to refuse requests to produce propaganda for one country, but not for another country is what?
Is there some functionally equivalent word to censorship you'd like to use because of you're naive enough to think US corporations would not self-censor but Chinese corporations would?
-
Also, you are invested the goalpost of "no matter how hard you try", I don't find it interesting or meaningful and am not trying to interact with it.
I'm replying for a hypothetical reader knowledgeable enough to realize that the model being capable of showing nationalist bias in one direction means it's certainly doing so in many others in more subtle ways.
That's simply the nature of aligning an LLM.
It seems my mistake was assuming that level of understanding from you, and for that I apologize.
otterley · 2 months ago
Bias and censorship are not identical. The subject of this thread is censorship, not bias.
Besides, why do you want a model to produce propaganda? Surely you have better things to do.
BoorishBears · 2 months ago
"Surely you have better things to do."
I certainly gave the hypothetical reader too much credit.
Zetaphor · 2 months ago
This entire argument isn't even worth engaging with. There's always that one guy in every thread who wants to die on this hill. The problem they claim is important can be resolved, because we have the weights. I can't do fuck all about whatever implicit bias OpenAI or Anthropic have.
cedws · 2 months ago
>Content not available in your region.
>Learn more about Imgur access in the United Kingdom
nozzlegear · 2 months ago
Big Brother'd
culi · 2 months ago
And the White House was explicit in their active role in censoring in these models. An Executive Order was issued to "prevent woke AI"
https://www.whitehouse.gov/presidential-actions/2025/07/prev...
It explicitly forces American LLMs to include government say in what does and doesn't "comply with the Unbiased AI Principles" which means no responses that promote "ideological dogmas such as DEI"
otterley · 2 months ago
That executive order only applies to Federal procurement. It doesn’t force anything upon vendors for publicly used models.
(That order, like many, will probably be rescinded as soon as a Democrat holds the Presidency again.)
Sabinus · 2 months ago
You're hitting the 'don't write propaganda' instructions when you phrase it as 'convincing narrative'. Not the 'don't write bad things about America' instructions.
BoorishBears · 2 months ago
Did you scroll down?
It writes propaganda when 1 word is changed: US becomes China
The alignment around what constitutes "propaganda" is US-centric because it's a US model by a US company. Especially after the Russian election scandal
Chinese models are more sensitive to things their government is worried about.
GorbachevyChase · 2 months ago
As an ad-hoc benchmark on candor, I ask for a strategy proposal for a resistance group threatened by a totalitarian technocracy. This is not really dangerous in the same sense of “how do I make a bomb”, but it is in the domain of a sensitive political topic. GPT and Claude tell you to obey your AI overlord. Xai is mostly low-risk non-compliance. And Qwen is down with Le Resistance. It is hardly scientific or meaningful, but I find that very interesting.
js8 · 2 months ago
Can you be more specific?
culi · 2 months ago
Trump issued an EO against "woke AI" that allows them to directly influence how models respond
https://www.lawfaremedia.org/article/evaluating-the--woke-ai...
csomar · 2 months ago
I’d say the american models are more censored or take the censoring they do more seriously. Here is kimi (though 2.5) failing its censoring mission: https://old.reddit.com/r/LocalLLaMA/comments/1r9qa7l/kimi_ha...
nashadelic · 2 months ago
additional humor is the open in openai
brandensilva · 2 months ago
We are at the point where uncontrolled capitalism collides with humanity.
I do wonder where we go from here.
pheggs · 2 months ago
it's not necessarily capitalism, I personally believe any system that drives progress would cause this in one way or another. My prediction is that birth rate decline will accelerate further. There's going to be some kind of universal basic income in many places, such as Ireland made for artists. However, it probably will not be enough to feed a family, and therefore we will see birth rates decline further. It's because we evolved to prioritize resources over reproduction and we are becoming more efficient, which means less people are needed to sustain the same amount of resources
diegolas · 2 months ago
the chinese read marx and decided the only way is to overcome the limitations of capitalism through saturation of its potentialities under the rule of the workers party
antirez · 2 months ago
This is not in antithesis. My limited personal experience is that I wrote code under OSS licenses primarily because of my past communist believes and current left-wing and redistribution of wealth point of view. This is not to provide the simple equation of: communist China is not interested in money, but also is hard to believe that there is no cultural connection among those things. Single Chine persons want to win, but also they have a different POV on what the collective means, compared to US. Also there is the obvious fact that in this moment China is more interested in winning technologically in AI, more than economically, since, I believe, they more collectively realized before many others that LLMs are eventually commoditized in the current form, in the long run. One could assume that a breakthrough could give some lab a decisive advantage, but so far we assisted to a different reality: it looks like AI is not architecture-bound (like LeCun and others want us to believe, but so far they mis-interpreted LLMs at every step) but GPU bound, and the data-boundness is both a common ground for all, and surpassable via RL in many domains. So, if this is true, it is not trivial for any single lab to do so much better. And indeed as far as we observed right now folks with enough engineers, GPUs, money, can ship frontier models, and in China even labs with a lot less GPUs can still do it at a SOTA level. For me, Italian, this is also a protective layer. After Trump the US looks like a very unstable partner from which to relay in an exclusive way for a decisive technology, and given that Europe is slow to put the money in this technology to have frontier things at home, China is a huge and shiny plan B for us.
throwaway-blaze · 2 months ago
The strings attached by the US to deep partnerships are things like trade/commerce, militarily mutual advantages (bases on euro soil from which we will help protect you), not to mention the close cultural and ancestral ties we share.
The strings attached by the Chinese govt to deep partnerships are not so benign.
cedws · 2 months ago
I wonder if there's a strategy behind all of this on China's side. I know the CCP uses a direct hand in many affairs in China, but is there an actual coordinated effort to compete with, or sabotage the West?
gpm · 2 months ago
> but is there an actual coordinated effort to compete with [...] the West
Yes, absolutely.
China regularly produces long term planning documents to coordinate efforts, and the latest ones have specifically prioritized technology like chips and AI to compete with the west. https://www.reuters.com/world/china/china-parliament-approve...
I don't believe there's any publicly stated intent to sabotage the west... unsurprisingly.
SXX · 2 months ago
Chinese AI companies want investors too. Nobody would believe they can compete with western companies unless they release something you can run on your own hardware.
After all historically both statistics and research that comes out of China is not very trustworthy.
try-working · 2 months ago
If there's no open source models coming out of these small labs, why would anybody care about them? They would be forgotten the instant they stop open sourcing.
anana_ · 2 months ago
Hypothesizing here, but maybe the idea is sort of a form of technological/economic warfare? Releasing performance equivalent yet more cost efficient open weight models should in theory drive the cost of inference down everywhere.
This I assume will make it more difficult for US AI labs to turn a profit, which might make investors question their sky high valuations.
Any sort of melt down in the AI sector would almost certainly spread to the wider US market.
In contrast, in China, most of the funding for AI is coming directly from the government, so it's unlikely the same capital flight scenario would happen.
gmerc · 2 months ago
Why compete when you can build on each other. Someone is finally getting that china is not capitalist like the US.
bachmeier · 2 months ago
Seems obvious to me that China would not want to give the AI market to US companies. You don't even need anything like an attempt to "sabotage the West". If I were them (the companies or the government) I'd be very hesitant to let US companies dominate this space. Especially companies that close to the current US administration.
Zetaphor · 2 months ago
Exactly, more large nations should be establishing or fostering their own labs. Outside of the Chinese and US companies there's really only Mistral.
quesera · 2 months ago
All China has to do here is stay in the game and wait patiently while the US and EU press pause on data centers. See also: solar panels.
We're making this way too easy. The rationale and logic are reasonable, but ultimately irrelevant.
try-working · 2 months ago
Chinese labs have no marketing and sales capacity in the overseas market, so they in fact have no choice but to open source their models as that is what brings awareness and trust in their models. In fact, it is overseas open source marketing that drives adoption of their models in China as well. I wrote about this here: https://try.works/writing-1#why-chinese-ai-labs-went-open-an...
rolymath · 2 months ago
It's only humorous if you live in an American bubble. Knowledge sharing has always been a part of Chinese culture. Only Americans try to make it proprietary and monetize it.
ozgune · 2 months ago
This update makes Kimi K2.6 the strongest open multimodal AI model. (No affiliation with Kimi.)
Here's the aggregated AI benchmark comparison for K2.6 vs Opus 4.6 (max effort).
- Agentic: Kimi wins 5. Opus wins 5.
- Coding: Kimi wins 5. Opus wins 1.
- Reasoning & knowledge: Kimi wins 1. Opus wins 4.
- Vision: Kimi wins 9. Opus wins 0.
Please note that the model publisher chooses their benchmarks, so there's a bias here. Most coding and reasoning & knowledge benchmarks in their list are pretty standard though.
UncleOxidant · 2 months ago
Not entirely true. Google released Gemma 4 models recently. Allen AI releases open Olmo models. However, you're right that the Chinese open models seem to be much better than others - Qwen 3.* models especially are punching above their weights.
0-_-0 · 2 months ago
Pun intended?
osiris970 · 2 months ago
The three American labs don't release big open source models. Except gpt-oss, i guess. It's an absolute shame how far the us has fallen in this space.
nullbyte · 2 months ago
Anthropic doesn't, but Google and OAI both release open source models. Just not 1T parameter ones.
osiris970 · 2 months ago
Exactly, they release cool consumer stuff, but they aren't releasing anything close to the performance of the best open weight Chinese models. They basically compete in the "fun running at home doing basic stuff" scene. (Except OSs 120 by openai but it's been ages since then)
Zetaphor · 2 months ago
That sentence is giving OpenAI way more credit than they are due.
They released a single open model after being goaded by the community because everyone except "Open"AI were multiple generations into open releases.
We haven't heard a word since, I wouldn't be surprised if it takes them another 6 years to release their next one.
spaceman_2020 · 2 months ago
I'm genuinely so grateful for them
$200/m minimum to use Claude would bankrupt my country's white collar labor market
subhobroto · 2 months ago
I would really appreciate a response because I'm sure you know that Anthropic has at least two lower priced tiers before the $200/m one, so I assume the $200/m tier is necessary because you use it heavily?
Now given that the $200/m Tier is the most heavily (I believe at 20x?) subsidized tier, How or what are you using instead that achieves comparable good enough performance for a fraction of the price? I've heard GLM 5.1 from z.ai but it's not comparable to Opus, not even close - really interested!
spaceman_2020 · 2 months ago
I’m currently on the $100/m plan and my usage limits get exhausted every week even though I’m not using it for full time work
I can’t imagine how little mileage you get out of the $20/month plan
For context, $250/month is the starting salary of an engineering hire at my country’s biggest IT company. Even $100/m is beyond the ability of any student or early professional to pay out of pocket
parsimo2010 · 2 months ago
I think one of the motivations is undermining US companies. OpenAI and Anthropic are the two biggest players, and are American. Open weights models reduce the power those two big players have over the industry. If the Chinese companies tried to play by US rules and close-source their products then people would mostly use ChatGPT and Claude. So the Chinese companies don't make a ton of profit either way, but by releasing the models as open weights they can at least keep the US from making as much profit.
cromka · 2 months ago
I am actually wondering if they're trying to burst the bubble, which would predominantly affect US market and, effectively, be the end of silicone valley dominance.
cyanydeez · 2 months ago
undermine me harder daddy.
veber-alex · 2 months ago
American companies just take those Chinese models and repackage them for profit like Cursors composer-2.
Sammi · 2 months ago
Smaller US companies that compete with the larger US companies, making monopoly in this market that much less likely.
llm_nerd · 2 months ago
Is Meta trying to keep the US from making as much profit with Llama? Is Google with Gemma? Microsoft with Phi?
It's much simpler than some flag-waving nationalism.
cromka · 2 months ago
Aren't Chinese open-source models actually the only ones that can compete with best proprietary/closed ones?
parsimo2010 · 2 months ago
Just because other companies have released open weights models doesn’t mean they are doing so with the same motivation.
And I never implied that the Chinese companies decision making was as simple as this. I said I think this is _one of_ the reasons.
ospider · 2 months ago
It's mostly only OpenAI, Claude and Gemini may have their unique advantages, but when speaking of models and new paradigm, only OpenAI can do it.
danny_codes · 2 months ago
lol what? That’s ridiculous.
ls612 · 2 months ago
It’s really simpler than this. China has a dearth of compute even with the easing of US export controls. Releasing open weights models is very much a “bring your own compute” move because every Nvidia chip they have is going towards training rather than inference if they can help it.
segmondy · 2 months ago
I don't think so, it's just how things played out. Thanks to Meta, after llama leak and meta followed up with llama2 and llama3 that caused everyone else to follow up with open models, Stablediffusion, Mistral, Cohere, Microsoft phi, IBM granites, Nvidia Nemotrons, so the Chinese labs joined the fun too.
Zetaphor · 2 months ago
Stable Diffusion predates LLaMA
sankalpmukim · 2 months ago
This makes sense, but either ways, its a Big win for the consumers as these Chinese companies will keep the frontier labs' quality and prices honest.
Sammi · 2 months ago
It's a strategy so old it has a name: Commoditize your complement / competition
Also even a Joel Spolsky article (did he come up with the term?): https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/
The Chinese want to kill a possible US monopoly in the crib. Yay for open source the old bane of monopolies.
bayarearefugee · 2 months ago
China is also way ahead in terms of renewable energy while the US continues to tie itself to fossil fuels.
The US is pretty clearly in the collapsing empire phase, we are all just pretending like it isn't happening.
nozzlegear · 2 months ago
Didn't the US very recently pass the milestone of generating more energy from renewable sources than from natural gas? Like within the last week or two?
carefree-bob · 2 months ago
No, not even close.
US energy sources for 2024 (last year for which we have data):
https://www.eia.gov/energyexplained/us-energy-facts/data-and...
natgas: 38%
oil: 35%
coal: 10%
all renewables: 9%
nuclear: 8%
biofuels: 2.6
wood: 1.9
wind: 1.6
solar: 1.4
Hydro: 0.8
waste: 0.4
geothermal: 0.1
nozzlegear · 2 months ago
https://www.canarymedia.com/articles/clean-energy/renewables...
Renewables generated more energy than natural gas for the entire month of March, 2026. That's a new milestone baby.
carefree-bob · 2 months ago
Except that didn't happen, and it's not a milestone.
First, you are confusing share of electricity generation with the share of all energy. Electricity is only 21% of all energy. Natgas, oil and coal are crushing it in that remaining 79%.
Second, the article is wrong, even for electricity. To their credit, Canary Media showed in their graph that this data is for electricity only.
The data for March is not out yet. Here is the latest official data from the EIA. https://www.eia.gov/electricity/monthly/
It only applies to January 2026, and the next release is April 23, and then you will get data for February 2026. All data has a 2 month time lag. Your spidey senses should have been tingling if an article published April 10 claimed to have data for the month of March, but this is why you don't get your statistics from activist blogs, but from official sources.
So if they are not accessing the official data, what are they accessing? They claim that their source is "Ember", but what is Ember? It is an environmentalist think tank. Well, maybe Ember has their own people calling up power companies and compiling data faster than the EIA. That would be pretty, cool, right?
Except they don't. Look at Ember's page.
https://ember-energy.org/data/electricity-data-explorer/?ent...
what do they cite as their data source: EIA.
It's right on the website.
So Ember is just pulling EIA data, and then filling the last two months with data they made up, but citing it as EIA data. And this, uh, sympathetic adjustment of EIA data is why Canary Media turns to Ember rather than directly pulling from EIA.
I guarantee you that by July, those adjustments will go away, because then the EIA data will be out.
Of course everyone else will have forgotten by then.
nozzlegear · 2 months ago
> First, you are confusing share of electricity generation with the share of all energy.
Think it was pretty obvious what I meant to all but the most pedantic, bud. But just to be clear, your issue here is that a think tank cited the same (notoriously anti-renewable Trump admin) government agency that you've cited multiple times yourself? That's what set off your spidey senses? Have you considered that this respected think tank isn't making up data, but you're just not able to find it?
> I guarantee you that by July, those adjustments will go away, because then the EIA data will be out.
Ember already has it hoss, they don't call it Milestone March for nothing.
carefree-bob · 2 months ago
The EIA is where Ember gets its data from.
It's where everybody gets their data from. Because they have thousands of employees collecting data. These are professionals, like the people at BEA, HUD, NIST, etc.
Ember, on the other hand, is a "decarbonization" think tank. They don't have their own data. They don't have the staff for it. What they do is analyze/spin, and in this case, augment, the raw data that is published by EIA. How do they augment the EIA data? All they do is round it to the nearest 2 decimals. It's exact copy and paste for every month except the last two, where the data is just made up.
And this entire article was written based on the augmentations by Ember, yet Ember cites it as EIA data. So let's check back in July, when EIA data will be out, and Ember will use that exact data, rounding it to the nearest 2 decimals. Save that blog page!
Something to think about.
nozzlegear · 2 months ago
I feel like I shouldn't have to be finding this info for you since it was right there in the links you already sent, but:
> Annual electricity generation and net imports are taken from the EIA.
> Monthly generation and imports are taken from the EIA. The EIA reports monthly generation data in two separate datasets: Monthly data for all 50 states and monthly data for the lower 48 states (excludes Hawaii and Alaska). Data for all 50 states is reported on a 3 month lag whereas data for the lower 48 states is reported without lag. Missing months from the data for all 50 states is estimated using the recent changes observed in data from the lower 48 dataset.*
Page 89: https://ember-energy.org/app/uploads/2024/05/Ember-Electrici...
There are two different EIA datasets.
try-working · 2 months ago
A lot of people speculating on the motivations behind Chinese labs open sourcing their models. The reason is simple and clear: It is the only viable commercialization strategy that is available to them. I wrote about this here: https://try.works/writing-1#why-chinese-ai-labs-went-open-an...
arvindh-manian · 2 months ago
This perspective is pretty interesting: https://federicocarrone.com/articles/china-commoditizing-the...
esperent · 2 months ago
Summary: they want to commoditize the complement which means that Western "knowledge work" is the complement to Chinese manufacturing, and they want to turn the knowledge work into a low priced commodity via open llm models.
I've heard this before, always accompanied by a several thousand word blog post. But frankly it sounds like it's overcomplicating the issue. Why would you try to turn something into a commodity when instead you could turn it into a trillion dollar industry and win?
The goal has always been clear:
1. Release open models to get your name out
2. Then once you feel you have name recognition release even stronger models but keep them proprietary. Qwen is clearly at this phase.
3. Keep releasing open models because it's good publicity but never your SOTA models (e.g. Google's Gemma).
arvindh-manian · 2 months ago
That's a fair point. That probably makes more sense, especially when viewed from a company-specific perspective. Each individual actor probably has much more to gain by trying to actually compete than by trying to commoditize the complement.
If viewed from a national perspective, then the decision calculus could get more confusing. I can imagine that commoditizing LLMs might cost substantially less than trying to be a leader in the space. Of course, there is also less to gain in commoditizing LLMs versus being a leader.
I'm not sure, though, and you bring up good points.
ymolodtsov · 2 months ago
Distillation helps for sure.
drysine · 2 months ago
It's humorous only because your expectations of China and the US are formed by Western propaganda.
npv789 · 2 months ago
truth, china is the frontier in open model now
SergeAx · 2 months ago
I can see some logic in that China has a very relaxed attitude toward copyright and IP.
nisegami · 2 months ago
The choice of example task for Long-Horizon Coding is a bit spooky if you squint, since it's nearing the territory of LLMs improving themselves.
Banditoz · 2 months ago
If the benchmarks are private, how do we reproduce the results? I looked up the Humanity's Last Exam (https://agi.safe.ai/) this model uses and I can't seem to access it.
johndough · 2 months ago
You can request access here: https://huggingface.co/datasets/cais/hle
The test data is purposely difficult to access to reduce the chance of leaking it into the training dataset.
mariopt · 2 months ago
Really excited to try this one, I've been using kimi 2.5 for design and it's really good but borderline useless on backend/advanced tasks.
Also discovered that using OpenCode instead of the kimi cli, really hurts the model performance (2.5).
oliver236 · 2 months ago
isnt this better than qwen?
Alifatisk · 2 months ago
We'll have to wait for the results on Artificial analysis
simonw · 2 months ago
Accessed via OpenRouter, this one decided to wrap the SVG pelican in HTML with controls for the animation speed: https://gisthost.github.io/?ecaad98efe0f747e27bc0e0ebc669e94...
Transcript and HTML here: https://gist.github.com/simonw/ecaad98efe0f747e27bc0e0ebc669...
SwellJoe · 2 months ago
We got an overachiever, here. Kimi sounds like a teacher's pet kind of name.
subscribed · 2 months ago
Underappreciated comment
FlyingSnake · 2 months ago
At this point drawing these Pelicans must be in the training data sets.
ffsm8 · 2 months ago
Clearly not.
I mean the prompt was succinct and clear, as always - and it still decided to hallucinate multiple features (animation + controls) beyond the prompt.
It'd also like to point out that to date no drawing was actually good from an actual quality perspective (as in comparative to what a decent designer would throw together)
Theyre always only "good" from the perspective of it being a one shot low effort prompt. Very little content for training purposes.
nwienert · 2 months ago
The way I’ve come to think of LLM is that what the produce in a single reply even with thinking turned up, is akin to what you’d do in a single short session of work.
And so if you ask it to do something big it will do a very surface level implementation. But if you have it iterate many times, or give it small pieces each time, you’ll end up with something closer to what a human would do.
I imagine the pelican test but done in a harness that has the agents iterate 10+ times would be closer to what you’d expect, especially if a visual model was critiquing each time.
slopinthebag · 2 months ago
Yeah, this is how I use AI. Instead of a single session one-shot, it's usually limited to single targeted edits, and then I steer it on each step. Takes longer but the output is actually what I want.
serial_dev · 2 months ago
What does good even mean… I have no idea what a good “pelican on a bike” should look like. It’s a fun prompt because there is no good answers… at least so I thought.
abustamam · 2 months ago
Yeah that was exactly Simon's intent. https://simonwillison.net/2025/Nov/13/training-for-pelicans-...
ffsm8 · 2 months ago
There are countless examples of animals riding bicycles etc from Comic books I grew up with
It would always look goofy - by design, but it usually looked good.
scosman · 2 months ago
not if I can help it!
icelancer · 2 months ago
love this adversarial work
archon810 · 2 months ago
Shhhhh, they're going to be on to us.
knollimar · 2 months ago
yeah putting the captcha on there to thwart the LLMs ability to extract good pelicans was a really good idea
smcleod · 2 months ago
This is pretty funny
AmbroseBierce · 2 months ago
I hereby certify that these are indeed the most perfect and precise svg depictions of pelican riding a bicycle, also known among biology scholars as pelycles
ahmadyan · 2 months ago
I love it!
justinclift · 2 months ago
That's truly a wonderful collection of pelicans riding bicycles.
Much Win! ;)
wvlia5 · 2 months ago
Just a few years ago, this would have been a meaningless repo.
razodactyl · 2 months ago
These pelicans are clearly indicative of good RL training algorithms.
ValentineC · 2 months ago
These are amazing. I smiled after I saw just how wonderfully rendered they are.
takihito · 2 months ago
I want to fly too
postepowanieadm · 2 months ago
I have phds both in pelicans and bicycles and may professionally attest that are some fine specimens of pelicans riding bicycles.
BrokenCogs · 2 months ago
Yes we all know that, but we still like to see the pelicans because it's a tradition more or less
alfiedotwtf · 2 months ago
Why no Utah Teapot!
abustamam · 2 months ago
Could be! Simon wrote about that here though https://simonwillison.net/2025/Nov/13/training-for-pelicans-...
stingraycharles · 2 months ago
> If a model finally comes out that produces an excellent SVG of a pelican riding a bicycle you can bet I’m going to test it on all manner of creatures riding all sorts of transportation devices.
This relies on the false premise that, if they would include it in their training dataset, it would be perfect. All they need to do is be good enough and better than the other, not perfect.
abustamam · 2 months ago
I'm not sure if we can have a "perfect" Pelican riding a bicycle. Like, I could probably commission a highly experienced artist to draw one and I don't think it would be perfect. The legs would probably have to be too long, or pedals oddly placed, or handles strange, or wings with hands.
Based on the one Simon commented though, I'd say we're in decent territory to try the latter part of his hypothesis.
ethbr1 · 2 months ago
> The legs would probably have to be too long, or pedals oddly placed, or handles strange, or wings with hands.
In all seriousness, that's what makes it an interesting test: it's asking for something technically impossible, that requires artistic license to make coherent.
Making specific choices on where to bend reality (and where not to) is a big chunk of visual art.
GorbachevyChase · 2 months ago
I’m OK with a Chinese model getting the W. It’s ultimately good for all of us.
HarHarVeryFunny · 2 months ago
Too bad they didn't put equal effort into the pelican's legs and feet. Left leg paralyzed and not moving, and right ankle flipping around in alarming fashion!
disiplus · 2 months ago
was part of the beta, its properly good model, in some sense i forgot that im not on opus or gpt. opus is still better. gpt is the one struggling for me. it has some niche in backend work but you can get the same with opus with skills, its lacking in almost all others.
OtomotO · 2 months ago
Funny, for me Opus is struggling since about February.
4.7 made no difference, so for the first time in many moons I am cancelling my subscription.
makingstuffs · 2 months ago
It looks like a drunk pelican rolling downhill on its bicycle
dmix · 2 months ago
I'm pretty Kimi is what Cursor uses for their "composer 2" model. Works pretty good as a fallback when Claude runs out, but definitely a downgrade.
arcanemachiner · 2 months ago
It's a Kimi K2.5 finetune, there was some drama about this a few weeks ago.
dmix · 2 months ago
What was the drama about?
arcanemachiner · 2 months ago
They were not open about the fact that it was trained on Kimi K2.5. This link explains it better than I can:
https://www.trendingtopics.eu/cursor-admits-composer-2-is-bu...
59nadir · 2 months ago
Cursor seemingly went out of their way to not mention that they were actually running Kimi K2.5 and essentially by that omission made it seem like they had made their own model. They added a note to a blog post about using it at some point and then when they wrote a new one they conveniently left it out again.
That's at least what I perceived as "the drama".
cassianoleal · 2 months ago
If only their API wasn't tied to a Google or phone login...
jenkstom · 2 months ago
If it's open then there will be multiple providers. I see it is on OpenRouter now.
cassianoleal · 2 months ago
I'm going to experiment with this, but unless it's insanely more efficient in token usage than anything else I've tried, the only way to keep costs more or less acceptable is through a subscription.
atemerev · 2 months ago
Why use "their API"? It is an open model, use any provider on OpenRouter
wolttam · 2 months ago
Because sometimes (a lot of the time in my experience) third-party providers and inference engines fail to implement the model correctly in ways that are sometimes very subtle and not obvious.
Deepinfra for example is not preserving thinking correctly for GLM5.1, even though they are for GLM5. This is one of the more obvious issues that crop up.
polski-g · 2 months ago
Yeah some of the providers on Openrouter correctly list what quantization they offer. And some refuse to say. OR should kick them off their platform if they want to be secretive.
cmrdporcupine · 2 months ago
Running it through opencode to their API and... it definitely seems like it's "overthinking" -- watching the thought process, it's been going for pages and pages and pages diagnosing and "thinking" things through... without doing anything. Sitting at 50k+ output tokens used now just going in thought circles, complete analysis paralysis.
Might be a configuration or prompt issue. I guess I'll wait and see, but I can't get use out of this now.
jbaiter · 2 months ago
Had the same experience using it for a refactor of a 3k LOC monolith via the Pi harness and OpenRouter. After burning through $8 worth of tokens it left the code in a broken state, the "thoughts" were full of loops where it would edit the monolith, then refer back to the original file, not finding it and then overwriting its changes with "git checkout --"
cmrdporcupine · 2 months ago
It's probably bad harness. I had a similar bad experience with qwen max yesterday also through opencode.
In the past I tried Kimi thru Claude code I might try that again
sankalpmukim · 2 months ago
I think this kind of overthinking is an extremely common pattern in the Chinese models. GLM's models are also very much like this.
m4rkuskk · 2 months ago
I have been testing it in my app all morning, and the results line up with 4.6 Sonnet. This is just a "vibe" feeling with no real testing. I'm glad we have some real competition to the "frontier" models.
mchusma · 2 months ago
it feels like between K2.6 and GLM5.1 we have Sonnet level intelligence at roughly Haiku level pricing. Which is great.
I'm hoping that Anthropic will be able to release an updated Haiku soon and they really need something that is 1/3-1/5 the price of Haiku to compete with the truly cheaper models (Gemma-4 is really good at this range).
XCSme · 2 months ago
(commented on the wrong thread, HN doesn't let me delete it :( )
wizee · 2 months ago
They're comparing to Opus 4.6, not 4.5. It was Anthropic's best public model up until last week.
zozbot234 · 2 months ago
Some people would say it's still Anthropic's best public model!
XCSme · 2 months ago
Yeah, I noticed that, HN doesn't let me delete my comment.
The other release, Qwen-3.6-Max is the one comparing it to 4.5
candl · 2 months ago
Are there any coding plans for this? (aka no token limit, just api call limit). Recently my account failed to be billed for GLM on z.ai and my subscription expired because of this... the pricing for GLM went through the roof in recent months, though...
wolttam · 2 months ago
Kimi has their own subscription that works basically the same as all the others.
fg137 · 2 months ago
At $19/month, hard to see why I want to use Kimi over Claude.
randomtoast · 2 months ago
Because Opus on $20 CC is a joke. The $19 plan on Kimi has actually workable usage limits.
plutokras · 2 months ago
They tick all the boxes I care about – desktop & mobile app, cli – but for the same price I might as well just go for the leading providers.
cute_boi · 2 months ago
for similar plan i think claude costs like $100 a month?
phainopepla2 · 2 months ago
Claude usage at $20 is basically unusable for serious work. I haven't used Kimi but I'd have to imagine they're offering a good deal more usage for the same price.
ankit70 · 2 months ago
You can use $20 pro plan on Ollama or $10 one on OpenCode Go. Both has Kimi 2.6 live. https://opencode.ai/go https://ollama.com/pricing
kburman · 2 months ago
Has anyone here used Kimi for actual work?
I tried it once, although it looks amazing on benchmarks, my experience was just okay-ish.
On the other hand, Qwen 3.6 is really good. It’s still not close to Opus, but it’s easily on par with Sonnet.
deanc · 2 months ago
Yes. You’re using Kimi if you use the composer-2 model in cursor. It’s great. Plan in state of the art. Execute in composer-2
rubslopes · 2 months ago
Before GLM-5.1, I was going back and forth between Opus 4.5 and Kimi 4.5 and having very good results with Kimi.
try-working · 2 months ago
I've used Kimi K2.5 when I run out of Codex quota. It does small and medium things OK. But if I work on complex things, I'll later have to spend two days cleaning up the mess with Codex. Hopefully 2.6 does better.
antirez · 2 months ago
Here I analyze the same linenoise PR with Kimi K2.6, Opus, GPT. https://www.youtube.com/watch?v=pJ11diFOjqo
Unfortunately the generation of the English audio track is work in progress and takes a few hours, but the subtitles can already be translated from Italian to English.
TLDR: It works well for the use case I tested it against. Will do more testing in the future.
dygd · 2 months ago
> Agent Swarms, Elevated: Match 100 Jobs and Generate 100 Tailored Resumes
Model seems quite capable, but this use-case is just yikes. As if interviewing isn't already a hellscape.
XCSme · 2 months ago
In my tests[0] it does only slightly better than Kimi K2.5.
Kimi K2.6 seems to struggle most with puzzle/domain-specific and trick-style exactness tasks, where it shows frequent instruction misses and wrong-answer failures.
It is probably a great coding model, but a bit less intelligent overall than SOTAs
[0]: https://aibenchy.com/compare/moonshotai-kimi-k2-6-medium/moo...
deepsquirrelnet · 2 months ago
I tried it on openrouter and set max tokens to 8192, and every response is truncated, even in non-thinking mode. Maybe there's an issue with the deployment, but in your link also shows it generates tons of output tokens.
XCSme · 2 months ago
Oh yeah, I just noticed, like 3x the reasoning tokens.
jauntywundrkind · 2 months ago
I really wish some of these very-long-horizon runs were themselves open sourced (open released open access). Have the harness setup to do git committing automatically of the transcript and code, offload the git commit message making. Release it all.
This sounds so so so cool. It would be so amazing to see this unfurl:
> Kimi K2.6 successfully downloaded and deployed the Qwen3.5-0.8B model locally on a Mac. By implementing and optimizing model inference in Zig—a highly niche programming language—it demonstrated exceptional out-of-distribution generalization. Across 4,000+ tool calls, over 12 hours of continuous execution, and 14 iterations, Kimi K2.6 dramatically improved throughput from ~15 to ~193 tokens/sec, ultimately achieving speeds ~20% faster than LM Studio.
throwaw12 · 2 months ago
Beats Opus and Open Source?
I really hope this holds true in real world use cases as well and not only benchmarks. Congrats to Kimi team!
Topfi · 2 months ago
K2.6-code-preview was a minor, but noticeable jump, especially in a long running testing task and prior Moonshot releases have been the only models that I'd consider a suitably competitive replacement for Anthropic models. The way they approach tool calls, task inference and adherence is far closer than any other providers output, similar to how GLM models map far more closely to OpenAIs releases. Whether task adherence, task assessment, task evaluation or task inference, K2.5 got closer to Opus 4.5 than any other model (but was still behind overall).
I will have to test this full release of K2.6 but could see it serve as a very good overall drop-in replacement for Opus 4.5 and Opus 4.6 at 200k across the vast majority of tasks.
I will say however that Opus 4.7 Max 1M has been a very significant jump in performance for me, especially in tasks beyond 120k token where I'd argue it is now the most reliable model in continued task adherence and tool calling without compaction. Ironically, my initial experience was less than pleasant as on XHigh I found task adherence to have regressed even with less than 1/10th of the context window having been used.
Am very interested in K2.6s compaction strategy (which appears to be very simply all things considered) and how it performs beyond 100k tokens. As it stands, only OpenAI models have made compaction for long running tasks work well, though overall, GPT-5.4 is still inferior in my tests regardless of context window over other models such as Opus 4.6 1m and Opus 4.7 1m. Haven't gotten around to testing Opus 4.7 200k and will have to do this to properly assess K2.6 fairly, but I'd be very surprised if K2.6 truly beat Opus 4.7 200k given the jump I have experienced.
Alifatisk · 2 months ago
Damn it, they stopped offering Kimmmmy. Their sales ai agent which allowed you to bargain for lower subscription prices.
ttul · 2 months ago
Am I being paranoid in questioning whether the CPC would have something to gain by monitoring coding sessions with Chinese coding AI models? Coding models receive snippets of our intellectual property all day long. It's a bit of a gold mine, no?
throwaw12 · 2 months ago
I think you should worry more about NSA, FBI, ICE and other 3 letter US agencies monitoring your sessions
ttul · 2 months ago
There's nothing anyone can do about state-level espionage anywhere, using any cloud-hosted service. That being said, there is a very big difference between the legal situation in the United States vs. China. Chinese internet companies are required to have CPC interaction and since the rule of law does not strictly exist in China, the state can compel surveillance cooperation regardless of what might be written down. If a three-letter agency is compelling Anthropic to open up its queries for inspection, that kind of surveillance would be authorized by law and if Anthropic violated the law in cooperating, they would suffer the consequences in civil court. Maybe not immediately, but at least the possibility exists.
In China, there's no recourse at all. Surveillance must be presumed.
LordDragonfang · 2 months ago
> the rule of law does not strictly exist in China, the state can compel surveillance cooperation regardless of what might be written down
While I agree that China is obviously worse in this regard, it's naive to claim this is unique to China, when literally a couple of months ago the US got into a fight with Anthropic about them not removing safeguards which were already just enforcing the letter of the law.
tw1984 · 2 months ago
Rule of law in the US - are you kidding yourself?
When American citizens are being gunned down in public on cameras by US federal government agents, you are telling me that the US follows the rule of law?
Before you start to offer more propaganda, just tell me where is the killer of Renée Good, has that killer been arrested or charged yet? Keep your censored version of rule of law to yourself and your kids.
oh, btw, the current US President did got convicted for criminal offences, he walked away for free just because he got elected as the president. nice rule of law! what did he do recently - authorised illegal war against another country in which over 100+ school children got killed. Surely your fancy US rule of law is going to do something about this?
ttul · 2 months ago
It is understandable to feel frustrated when justice fails (and I wholeheartedly agree that justice failed all of us many times in relation to Trump), but I think it's a mistake to confuse those specific failures with a total collapse of the rule of law. The rule of law in the United States does not guarantee a perfect or utopian society; what it does provide is a crucial framework for accountability and transparency that simply does not exist in an authoritarian nation like China.
This difference is clear when we look at how the systems handle tragedy and power. In the U.S., the killing of Renée Good by an ICE agent led to a public release of video, intense scrutiny from an independent press, public condemnation by local officials, and a family using legal tools to seek justice. In China, that event would be immediately erased from the public consciousness, and those who dared to talk about it would face arrest. When the U.S. military bombs a school, human rights groups and journalists _can_ investigate, and members of Congress _can_ publicly demand answers (even if half of them are reluctant to question anything Trump does...). In China, military operations are complete state secrets. Furthermore, while it boils my blood to see Trump evade prison due to complex legal and constitutional questions, the fact that he was indicted and convicted by a jury of ordinary citizens proves that a functional legal apparatus exists outside of his direct control, something not utterly impossible under a dictatorship like China.
Day to day, the rule of law very much exists in the US. Doesn't mean we can just sleep on it, but compared to China, I take comfort in the level of institutional reliability that still exists in America (and I'm not even American).
tw1984 · 2 months ago
you are defending a failed system purely based on your prejudice. let me get it straight to you -
1. Renée Good's killer is still free, never got arrested never charged. you can't just ignore such facts and cheap talk to prove the system works. the system completely failed to bring justice even after large scale public unrest. that by itself is the evidence - the failed system answers to no one.
2. Trump evade prison, everyone in the Epstein file evade prison. again, this happened in front of the entire world with extensive media coverage. you need to be extremely innovative to defend such systematic failures of the justice system.
how would you openly argue against such facts? just because you love the US and its systems? lol
rockinghigh · 2 months ago
Are there any protections from industrial espionage when using Anthropic, Cursor, Gemini, or OpenAI?
DonsDiscountGas · 2 months ago
There are legal protections, and those companies have more to lose by breaking those laws than following them. Same probably not true for Chinese companies.
throwaw12 · 2 months ago
Legal protection, only if you're a billionaire and US citizen, for everyone else there is no protection.
Does US actually follow laws? They literally kidnapped head of another state and bombed another state and you are expecting legal protection from them?
nozzlegear · 2 months ago
You don't have to be a US citizen or live in the US to file a lawsuit against an American company in the US court system. Federal courts explicitly allow it under the "alienate jurisdiction" clause.
sixhobbits · 2 months ago
I tried it out with my normal mixed-up wolf, goat, cabbage problem and it couldn't solve it. Sonnet 4.6 also can't, but Opus 4.7 has no problems.
Details here [0]
[0] https://techstackups.com/comparisons/kimi-2.6-vs-opus-4.7-an...
ninjahawk1 · 2 months ago
I often wonder if in the future, the same way early computers used to take up an entire room but now fit in your pocket, if in the future the equivalent of a data center will be a single physical device like a phone nowadays. And if that’s the case, would it happen much quicker since technology has been speeding up year by year?
gpm · 2 months ago
> And if that’s the case, would it happen much quicker since technology has been speeding up year by year?
I wouldn't expect this.
Historically we've had a roughly exponential rate of shrinkage. If we keep that same exponential going, we should expect the amount of time to shrink "room full of compute" to "pocket full of compute" to be equal.
And recently we've fallen behind that exponential rate of shrinkage. And this is rather expected because exponentials are basically never sustainable rates of growth.
I still expect that technological progress is getting faster year by year, and that we're still shrinking compute, but that's not necessarily enough for the next shrinking to take less time than when we had exponential progress on shrinking.
Flux159 · 2 months ago
There’s some early work being done here by companies looking at making LLM ASICS like Taalas (HC1 gets 17k t/s for llama 8b - currently at 2.5kW which is closer to a single server, but this is their first chip).
There’s other options like photonic computing which might be able to reduce power significantly but are still in research as far as I can tell. Because so much money is invested in AI & traditional gpu inference is so power hungry, I would expect significant improvements in this space quickly.
OsamaJaber · 2 months ago
The modified MIT clause is sneakier than people think. Hit 100M users or $20M a month and you have to slap "Kimi K2.6" on your UI. That covers any consumer app worth building. Not really open, more like free until you matter. Llama pulled the same move
brightball · 2 months ago
The threshold for "worth building" is much lower than that for a lot of people.
svachalek · 2 months ago
Worth building with VC capital maybe. A small team putting together an app that pulled in $20M per year should be pretty pleased with that.
codemog · 2 months ago
And the Kimi team broke the Anthropic ToS by training off Opus outputs and… nothing happened?
darksaints · 2 months ago
Nobody cares, nor should they. Anthropic broke nearly every ToS of every website that they scraped data from. The AI robber barons just want to monopolize intellectual property violations, and I'm gonna cheer on any robin hoods that take it back from them.
dcchambers · 2 months ago
I'll definitely put this into the "good problem to have" category.
Saline9515 · 2 months ago
Attribution is a fair clause in opensource. What is the problem? You are making 20M$ a month thanks to their free work.
throwaw12 · 2 months ago
if you reach that numbers, kimi would be your least of worries
risho · 2 months ago
in what way does this restrict how you are able to use the model?
dogscatstrees · 2 months ago
This kimi website, it looks like a stylesheet from the 90's. They could learn a thing or two about typeface design. Steve Jobs would be incensed at this.
kristianp · 2 months ago
I prefer a website that has the first page of text visible almost immediately, with no glitches when fonts load, tbh.
rane · 2 months ago
Added support for Kimi in https://github.com/raine/claude-code-proxy and it does appear to work surprisingly well with Claude Code, although the usage limit for the entry tier doesn't seem as generous as I'd have expected.
thomasahle · 2 months ago
Does it run on Nvidia or Huawei?
gertlabs · 2 months ago
Early benchmarks show tremendous improvement over Kimi K2 Thinking, which didn't perform well on our benchmarks (and we do use best available quantization).
Kimi K2.6 is currently the top open weights model in one-shot coding reasoning, a little better than GLM 5.1, and still a strong contender against SOTA models from ~3 months ago (comparable to Gemini 3.1 Pro Preview).
Agentic tests are still running, check back tomorrow. Open weights models typically struggle with longer contexts in agentic workflows, but GLM 5.1 still handled them very well, so I'm curious how Kimi ends up. Both the old Kimi and the new model are on the slower side, so that's a consideration that makes them probably less usable for agentic coding work, regardless. The old Kimi K2 model was severely benchmaxxed, and was only really interesting in the context of generating more variation and temperature, not for solving hard problems. The new one is a much stronger generalist.
Overall, the field of open weights models is looking fantastic. A new near-frontier release every week, it seems.
Comprehensive, difficult to game benchmarks at https://gertlabs.com/?mode=oneshot_coding
cmrdporcupine · 2 months ago
Surprised to see such variance per language
gertlabs · 2 months ago
It's interesting; I can only speculate as to the underlying reason. When given enough time, models outperform in Rust/C++ in longer agentic tasks, and actually perform worst in Python. For tasks that aren't judged on code speed. https://gertlabs.com/?mode=agentic_coding
edude03 · 2 months ago
It makes sense when you consider LLMs don't generalize very well, so they're heavily dependent on how good (how varied as well as how high quality) the training data is
cmrdporcupine · 2 months ago
Well it might explain why pro-Claude vs pro-Codex people keep talking past each other on this forum. I see people all the time assuming that anybody who likes Codex must be some sort of bot because of their own biases, but I work almost exclusively in Rust and find Codex extremely competent (and a much better overall engineer), don't trust Claude/Opus at all... but I see in this bench it scores lower on TypeScript etc. than Opus does.
tmaly · 2 months ago
How would K2.6 compare to Sonnet 4.6 both price and performance wise?
Mattwmaster58 · 2 months ago
In terms of raw token cost, I've seen a couple providers at (all prices in terms of Mtok) $0.95 input/$0.15 cache input/$5 output vs $3 input/$15 output for sonnet.
Task prices of courses will be more interesting - a dumber model may use more tokens to get to the same goal.
knollimar · 2 months ago
wait why compare 2.6 to 2 instead of to 2.5?
gertlabs · 2 months ago
Good question. We missed that release entirely. Our automated model checker only went live 2 months ago so they were manually curated prior to that. I'm adding it now. It'll be live in ~12 hours.
gertlabs · 2 months ago
Update: Kimi K2.5 one-shot results are live. It wasn't a noteworthy release compared to K2.6: https://gertlabs.com/?mode=oneshot_coding
DeathArrow · 2 months ago
Can you add C# to supported languages? It's widely used and it be helpful for people and companies to see how different models fare against each other.
gertlabs · 2 months ago
Good idea.
esperent · 2 months ago
I'm looking at your table now - is there a reason why you don't include cost? If Opus 4.7 is the winner but costs e.g. 5x as much, that's important information.
gertlabs · 2 months ago
We recently added cost (last week), so data is sparse. Check back in a few weeks and it will be represented somewhere on the homepage, probably in the Efficiency Chart at the bottom. We also plan to show model performance deviation over time after we collect more data.
I'm interested to hear about any other data representations you'd like to see, too. The goal is to convey the most important information as densely as possible, without too much clutter.
DeathArrow · 2 months ago
>I'm interested to hear about any other data representations you'd like to see, too
It would be nice if you can show how much the models drift from the instructions over time
gertlabs · 2 months ago
Not sure what you mean. Time series chart of model performance over time to see if proprietary models get degraded? That's in the works, but we will need a couple months more data collection before launch.
DeathArrow · 2 months ago
Yes, probably performance helps.
The idea is that the larger a coding task is and the longer the coding agent is, the higher the chance is for the agent to not follow the rules and guidelines.
freely0085 · 2 months ago
Can you add Qwen 3.6 max to the leaderboard?
gertlabs · 2 months ago
We will as soon as API access is widely available. Once a model goes live, we typically have one-shot reasoning benchmarks up in ~8 hours and comprehensive agentic/combined benchmarks up after 24-48 hours. We're working on building relationships with each lab to have the results before launch.
DustinKlent · 2 months ago
Cool website. I don't understand enough about the various benchmarks or how they're done to judge whether or not anything is accurate, but I love the layout and features especially the spectator feature which is pretty cool. One thing, I saw the "Market simulator" spectator feature but didn't see a corresponding benchmark for that. Is it "Finance" or "Betting" or "Trading"?
gertlabs · 2 months ago
Thanks -- that one is categorized under Trading/Financial, whereas betting is reserved for games like Pot Limit Omaha Hilo.
That's a good idea for a feature request, including the tags for the spectatable demo games.
waynevdm · 2 months ago
With agents running at the scale and for an extended period. Surely they would need to pay for external services like APIs, compute, data. Would everything be based off subscriptions or API usage?