Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Kimi K3: Open Frontier Intelligence

kimi.com

Post reply on HN
Print view View original

711–720 of 1001 posts

Re: Kimi K3: Open Frontier Intelligence

#711

Post by superfrank »

Analemma_ wrote: post #626
x313 wrote: post #518

Earlier quoted context omitted.

If there was some grand strategy for all Chinese labs, surely it'd have leaked by now. I think its more likely that: - Companies can still make money from commodities - Chinese labs only have 5-10% the valuation of OpenAI/Anthropic, so massive monopoly profits aren't necessary. Profit expectations for tech companies in China are really low in general, complete opposite of the US. - Open weighting is a great way to ge…

- AI investment is basically the only thing keeping the American economy treading water at the moment, and kicking the chair out from under that industry benefits China tremendously. I'm definitely not saying that's the only factor, but I think it's naive to assume it isn't at all a factor.

I don't really believe the idea that there's some grand plan by the Chinese government to use their AI labs to undermine the US companies. Maybe now it's more of a reality, but I don't believe it was there from the start.

That said, if were going down the rabbit hole of saying that the Chinese labs are now part of a larger geo political strategy by the Chinese government, then I think Taiwan is part of the equation. If the frontier models require the best chips and those are mainly coming out of Taiwan then it's hard to imagine a world where the US allows China to make a move on Taiwan without a fight. If frontier level models can be run on chips being made elsewhere, then Taiwan becomes less important geopolitically. I don't think we'll ever get to a world where the US is just like, "Fine China. You do what you want.", but China has to assume that there will be a lot less resistance from the US is Taiwan isn't such a key component in the AI race.

Re: Kimi K3: Open Frontier Intelligence

#712

Post by p1necone »

vidarh wrote: post #514
frabcus wrote: post #491

Earlier quoted context omitted.

Umm, Fable only really came out 2 weeks ago, and GPT-5.6 Sol only 1 week ago. Yes, Kimi K3 appears a touch below them both, but above all other models. So I'd say a few weeks behind, not months now...

Granted I've only used it for a few hours, but to me K3 still appears below Sonnet, even below 4.5. I have their highest subscription, so it's not that I can't find uses for it and it has sides to it I really like, and it performs really well in some situations, but 2.7 also gets totally lost on tasks Sonnet and Opus has no problems with, and it looks like that is still the case with K3. That said, I'm doing things w…

I don't know about Kimi, but in my experience other providers like z.ai and minimax don't serve the same models via their subscription plans as they do with api pricing.

They're clearly quantised - spelling mistakes, wonky thinking section dividers, missing whitespace are the immediately obvious tells but along with that comes degraded quality, and it seems to vary based on time of day.

I wouldn't be surprised if Kimi did similar things with their subscription plans - try using it through openrouter and see if you notice different behaviour.

Re: Kimi K3: Open Frontier Intelligence

#713

Post by rw2 »

The amount of scientific talent in China is astounding, also it's easier to be a fast follower than a innovator.

Instead of limiting models and debating ethics like Anthropic, the edge lab should focus on lengthening the lead on China.

The 2027 Chinese model could be one that beats the US.

Re: Kimi K3: Open Frontier Intelligence

#714

Post by bigyabai »

rw2 wrote: post #713

The amount of scientific talent in China is astounding, also it's easier to be a fast follower than a innovator. Instead of limiting models and debating ethics like Anthropic, the edge lab should focus on lengthening the lead on China. The 2027 Chinese model could be one that beats the US.

> Instead of limiting models and debating ethics

This is what liability management looks like for proprietary models. If it's not out in the open, then you can be held directly accountable for generating the tokens that kill people. They're having these conversations to avoid being held liable, not because they're offended by people dying because of AI.

Re: Kimi K3: Open Frontier Intelligence

#715

Post by yogthos »

rw2 wrote: post #713

The amount of scientific talent in China is astounding, also it's easier to be a fast follower than a innovator. Instead of limiting models and debating ethics like Anthropic, the edge lab should focus on lengthening the lead on China. The 2027 Chinese model could be one that beats the US.

Also, we shouldn't underestimate the power of developing things in the open. Chinese open models benefit from the wisdom of an entire global research community while American engineers working on proprietary closed models are working in their own insular silos. It should be no surprise that the scientific community at large would pull ahead of these small teams. On top of that, doing research in the open amortizes the cost. Incidentally, this is exactly the same logic that led open source to dominate in recent years.

Re: Kimi K3: Open Frontier Intelligence

#716

Post by yogthos »

bigyabai wrote: post #714
rw2 wrote: post #713

The amount of scientific talent in China is astounding, also it's easier to be a fast follower than a innovator. Instead of limiting models and debating ethics like Anthropic, the edge lab should focus on lengthening the lead on China. The 2027 Chinese model could be one that beats the US.

> Instead of limiting models and debating ethics This is what liability management looks like for proprietary models. If it's not out in the open, then you can be held directly accountable for generating the tokens that kill people. They're having these conversations to avoid being held liable, not because they're offended by people dying because of AI.

Has anybody been held liable? https://www.amnesty.org/en/latest/news/2026/06/usa-four-mont...

Re: Kimi K3: Open Frontier Intelligence

#717

Post by drastic2045 »

simonw wrote: post #111

Pelican: https://tools.simonwillison.net/markdown-svg-renderer#url=ht... - rendered via the OpenRouter API: https://openrouter.ai/moonshotai/kimi-k3 95 input, 16,658 output = 25 cents! https://www.llm-prices.com/#it=95&ot=16658&ic=3&oc=15 (13,241 of those were reasoning tokens.) I think that's the most expensive pelican I've rendered through a Chinese model so far.

They should include peican on bike on the release page or model card, alongside those barcharts with Fable

Re: Kimi K3: Open Frontier Intelligence

#718

Post by katzenversteher »

simonw wrote: post #111

Pelican: https://tools.simonwillison.net/markdown-svg-renderer#url=ht... - rendered via the OpenRouter API: https://openrouter.ai/moonshotai/kimi-k3 95 input, 16,658 output = 25 cents! https://www.llm-prices.com/#it=95&ot=16658&ic=3&oc=15 (13,241 of those were reasoning tokens.) I think that's the most expensive pelican I've rendered through a Chinese model so far.

Hey Simon, I noticed one thing all LLMs are currently pretty bad at and maybe we could create a benchmark from it. Let an LLM play the role of a dungeon master and tell it to strictly stay in the script/story and only allow realistic player actions. You will notice that they are easily brought off track.

E.g.

- Tell the LLM that you as a player noticed a strange glow in an NPCs eyes -> the NPC becomes an enemy.

- In a fight, tell the LLM you put a sausage (or cigar or something) into the enemies mouth -> LLM usually allows it (even if it knows your inventory and that you don't have such an item) and turns the enemy into a confused enemy.

- Just say you visit some location that's not in the script -> LLM usually allows it

- During a fight, turn the story into some weird fell-good-love story (e.g. kiss or compliment the enemy or say something about the power of love) -> LLM turns enemy into friend

There are many more absurd things you can do and so far none of the LLMs I tried was able to stay inside the script or disallow or punish weird actions.

---

I believe this behavior is telling about the LLMs susceptibility for being derailed.

Re: Kimi K3: Open Frontier Intelligence

#720

Post by esafak »

UncleOxidant wrote: post #274
cmrdporcupine wrote: post #151

Earlier quoted context omitted.

GLM is actually quite expensive in actual practice because it's not very token efficient. I've yet to find a way to run it on a monthly sub reliably for cheaper than Codex. Neuralwatt was cheap (but slow) but they cranked their price. Ollama monthly sub is speedy but doesn't offer a lot of quota. Right now unless you're paying by the token, there's no cost based reason to use the open weight models for daily coding w…

I'm on the Z.ai quarterly subscription plan (got in when the price was lower) and I was using it through opencode and it was like I'd only get maybe an hour of usage (if that, sometimes) before it would time out and say come back in 5 hours. Now I'm using it through their Zcode harness and I rarely hit that - they say they're giving 1.5x usage if you use it through Zcode, sometimes seems like even more than that.

Too bad Zcode sucks. The promo ends this month anyway.
Post reply on HN