Top
Best
New

Posted by vincent_s 7 hours ago

Kimi K3 is now live(www.kimi.com)
732 points | 425 commentspage 2
h2aichat 5 hours ago|
Working with chinese models is giving me a fullfilment sensation. I think that I have enough quality for the work that I need to do and lots of extra tokens to work with. With Claude and ChatGPT I reach the limits fairly easy, but not with OpenCode Go. So I will use Claude once in a while for difficult tasks to see how much better it still is (but use Chinese on a daily basis)
cg5280 2 hours ago|
I have been using Deepseek V4 Pro for personal projects and it has been great. I think the $20/mo GPT plan is still the strongest value, but only because you don’t have to pay API prices for tokens.
sebmellen 3 hours ago||
My testing prompt for these models is by no means objective or repeatable (like the pelican) but it's a nice test of curiosity:

> Impress me with a 1 page html file

Result: https://ydaurtg3fdwhq.kimi.page/

Came out looking pretty cool! By contrast, Fable produced a moderately more interesting "live observatory" of the solar system.

zorked 29 minutes ago||
I asked the same and got something vaguely similar. Then I asked for a demoscene-inspired demo in a non-traditional setting.

https://recherche-demo.kimi.page

nikcub 1 hour ago|||
Those thin capitalized eyebrows are becoming like the emdashes of visual design
tomashubelbauer 2 hours ago|||
This is a cool idea. I know I'd rather see this comment on every model release than the pelican.
gentlewater 2 hours ago||
Hah, that is indeed a pretty cool result.
buildbot 6 hours ago||
Amazing to see an open source model already nearing the benchmarks of Fable and GPT 5.6 Sol!

Also very cool to see LatentMoE being picked up by more models (https://arxiv.org/abs/2601.18089)

kroaton 5 hours ago||
It also goes to show that Fable/Sol must be 4-5T in size.
NoImmatureAdHom 6 hours ago||
Surely it's only open weights?
stefan_ 5 hours ago||
It's not even that right now.
buildbot 5 hours ago||
And they have since removed that language…
z4y5f3 3 hours ago||
They will release the weights by 7/27 along with support in vLLM. Stop second guessing. Source: their blog post https://mp.weixin.qq.com/s/V4xhEIy8xDXSMDPrPkmUAQ
buildbot 3 hours ago||
Thanks for the link. No need to be so aggressive. The blog with that detail was not live before; and they removed that language from the original link in this post.
swimwiththebeat 2 hours ago||
Did anyone see on the blog post[0] that it was able to code up an entire GPU compiler from scratch? It looks like it even outperformed triton on some GPU kernels. That just seems insane to me.

Wonder if they’ll open-source this and show how many tokens it cost.

[0] https://www.kimi.com/blog/kimi-k3

XCSme 4 hours ago||
I finished benchmarking[0] it, but it was not fun, it only supports (max) reasoning and the model is quite slow. Apart from a few requests timing out, it also has some issues with tool calling/response format schemas (Moonshot rejected tools.function.parameters with anyOf schema).

It also, for some reason failed to generate either of the 2 coding demos (hamster svg and solar system css animation).

Intelligence-wise, it's between GPT-5.6 Terra and GPT-5.6 Sol. It's ~30% better than Kimi K2.6, but a lot slower and more expensive.

[0]: https://aibenchy.com/compare/moonshotai-kimi-k3-max/moonshot...

XCSme 4 hours ago|
Just saw the logs, coding demos failed due to the 5 minute/task timeout. I have increased it and retesting it now.

EDIT: With 10 minutes timeout, the CSS task completed, but the SVG generation task still timed out. Trying again with 30 minutes timeout...

EDIT2: It completed (now in only ~9 minutes). It's one of the best hamsters[0].

[0]: https://aibenchy.com/compare/moonshotai-kimi-k3-max/moonshot...

msdz 6 hours ago||
> We also further increased the sparsity of the Mixture of Experts (MoE): with the Stable LatentMoE framework, the model efficiently activates 16 out of 896 experts. Together with improvements in training methodology and data recipes, these structural advances give K3 roughly 2.5x the overall scaling efficiency of K2, converting compute into capability more effectively.

Assuming experts are uniformly distributed (I’m really not that familiar with the deep details there), that’s 2800/896*16 = 50 billion active parameters just for the active/expert part. Wild stuff, and I’m glad there’s at least some companies still publishing (and pushing, for open-weight models) total parameter count.

And: It sounds very believable that this would result in efficiency gains wrt. to compute necessary for “good”-quality inference. Does anyone know whether there currently even are any SOTA or near-SOTA models that are dense still?

7734128 6 hours ago||
No, you can't divide the entire size by the expert count. A lot of weights are constant for all tokens, so total active count is ((2800-(shared)/896)*16 + (shared))
msdz 6 hours ago||
TIL, that makes a lot of sense, and thanks for the correction.
HarHarVeryFunny 5 hours ago||
Just to add to that, a Transformer block consists of an attention part followed by a feed forward part. MoE only modifies the feed forward part (which basically contains declarative knowledge getting injected into the residual stream).
Aeolun 6 hours ago||
2.5x the scaling efficiency, so 4 times the price? What is happening here? Did the subsidies dry up with the discrepancy between chinese and US models?
pixl97 6 hours ago|||
Scaling efficiency simply means if you took the first small model and scaled it up to the big model it would take 2.5x the resources to run. Not the that larger model is going to be any cheaper.

Kind of like scaling your personal automobile to the weight of a semi, the semi is still going to be far more efficient in moving cargo, not that the semi will cost the same to operate as the original car.

petu 6 hours ago|||
It's also 2.8x parameter count (1T -> 2.8T), likely higher activation per token (50B?).
InsideOutSanta 2 hours ago||
The blog post is now online:

https://www.kimi.com/blog/kimi-k3

- The blog post is explicitly saying that the model is open; that language was removed from the previously shared link

- It shows benchmarks

I've been playing around with it for the past few hours, and I think it's an amazing model. I'm not sure I could tell the difference between this and Fable in a blind test. The quota in the $100 Kimi Coding plan seems to roughly align with what I get from the $200 Anthropic plan when I primarily use Fable.

cesarvarela 19 minutes ago|
[dead]
blovescoffee 6 hours ago||
Excited for the deepseek release this week (or at least they announced they'd release this week). Hopefully they also push even closer to SOTA.
bayesianbot 6 hours ago||
That is exciting!

I don't understand how DeepSeek can be so cheap with their cache pricing - ~0.003 usd / 1Mtok. 100x less than Kimi K3, or similar numbers against pretty much any other decently sized model to my knowledge. I've been using it whenever possible as even longer agent sessions cost few cents.

sudosysgen 6 hours ago|||
If you read DeepSeek's papers, you'll find a litany of architectural features that allow for a greatly reduced cache hit price by shrinking the size of the KV-cache.
yfontana 6 hours ago||
How come no other big model seems to be able to deliver the same type of extremely low cache cost though, if their techniques are public?
jboss10 5 hours ago|||
I think the "architectural features" are part of the model, not the kv cache. So implementing it would be difficult and expensive.
petu 5 hours ago||||
Deepseek V4 paper is just ~three months old
sudosysgen 5 hours ago|||
Many of these techniques haven't been published very long ago - it often takes a good 6-8 months for techniques to percolate. But also, they come at a complexity cost and, seemingly, also at a stability cost.
hnfong 2 hours ago||
Also potentially a performance (in terms of output quality) cost. DeepSeek is cheap on a per token basis but lags behind in the benchmarks, perhaps it was a calculated tradeoff.
hack1312 6 hours ago|||
What provider are you using?
bayesianbot 6 hours ago||
DeepSeek's own API
greyb 6 hours ago||
Any way to avoid China sales tax or is that just the cost of doing business?
NortySpock 6 hours ago|||
https://openrouter.ai/deepseek/deepseek-v4-pro#providers

Look through the provider list for a company you are willing to do business with?

anigbrowl 2 hours ago||||
Good grief, the sales tax is only 6% on a service that's already extremely affordable.
verdverm 5 hours ago|||
Fireworks.ai
kamranjon 6 hours ago|||
Where did you hear about the deepseek release? Would love to follow the same source.
benjiro29 4 hours ago|||
> Where did you hear about the deepseek release?

* Tons of gray testing going on for the last 2+ weeks (people at random getting the new v4 model for a while before its removed again).

* It also DeepSeek their 3th birthday this Friday.

* The its been almost 3 months from the v4 DeepSeek release, and the model everybody have been using, was not post-trained. That is what they have been doing during this time.

People trying out the new DSv4 via the web chat with quick game creation tests. People pulling out stuff like Stellaris clones etc.

https://cct124.github.io/HORIZON6_DEMO/

https://www.showyourcode.app/zh/share/pmpwkamrnai2ue

The Battlefront like game is impressive. Sure, the soldiers are backwards and the graphics are still kind of basic. But the entire movement system (run/walk/crouch/jump), gun mechanics, grenades, capture points, AI fighting / capturing back, etc ... Ended up playing it way too darn long lol The text is in mandarin but its not too hard to figure out the menu. Sniper is OP ;)

The Horizon 6 game has everywhere mesh colliders, shows when you off track dirt being kicked up, etc ... In general, both example are very well polished minus the reverse soldiers issue.

And the price is supposed to stay the same (beyond the doubling during Chinese workhours), because everybody got that update.

blovescoffee 6 hours ago|||
They emailed current paying users of the api (or at least that’s how I got updated).
surgical_fire 5 hours ago||
Ohh I didn't know about it. Finally something to be excited about.
xyzsparetimexyz 6 hours ago||
Any updated Pareto frontier graphs? https://paraplouis.github.io/llm-pareto-frontier/ is quite out of date now.
tao_oat 6 hours ago||
I generally rely on LMArena for this: https://arena.ai/leaderboard/code/webdev/pareto

But it does take some days after model release before they collect enough data.

mdasen 3 hours ago|||
LMArena's "code" leaderboard is really skewed since it's a front-end JS code and design leaderboard. It generates a demo app with two models and then asks "do you prefer A or B". People can look at the code, but most of the time it's just going to be which one looks nicer.

Models that people like the design aesthetic of (Claude, GLM) tend to do better in LMArena than they do on other benchmarks. Design matters, but you look at a model like GPT-5.5 and it's behind Kimi K2.6, Sonnet 4.6, Qwen3.7 Max, and GLM-5.1 on LMArena's code leaderboard. Then you look at benchmarks like DeepSWE and GPT-5.5 blows them out of the water with only Fable and GPT-5.6 beating it.

I'm not saying that the LMArena leaderboard isn't useful, but I'm not sure how much weight I'd give it as a "code" leaderboard. I think often times it's a design comparison of simple front-end React apps rather than a coding comparison. GLM-5.2 is a very good model, but when you look at DeepSWE or Terminal-Bench v2, GPT-5.5 is well ahead.

Dibes 4 hours ago|||
Odd that open AI models aren't on that graph but are on the rankings! Must be a data lag issue?
Bromeo 6 hours ago|||
openrouter->rankings shows a pareto frontier. https://openrouter.ai/rankings#benchmarks
1899-12-30 4 hours ago||
you can get a rough version via artificialanalysis's cost per task https://artificialanalysis.ai/?cost=intelligence-vs-cost-per...
esher 7 hours ago|
Half kidding feature request for HN: Mark all AI related posts so I can filter them out, when I need a pause.
lfx 6 hours ago||
Here you go https://tools.simonwillison.net/hacker-news-filtered
mrtksn 6 hours ago|||
This post is at the top when filtered against AI :) Maybe it should use llm based filters to understand if the post is about AI and filter it out?
cyanydeez 6 hours ago||
Us the AI to build the bubble against the AI, because everyone knows AI is the AI of the AI.
postalcoder 6 hours ago||||
I'll see your simonw tool and raise you one that actually works: https://hcker.news/?view=frontpage&ai=exclude

I's not just matching against titles. Ironically, I have an agent running daily scans, reading the contents of the top 200 stories of the day. It auto screens high-confidence ones and I make judgement calls on like 10-20 of them per day.

epihelix 6 hours ago||
Right now, that site doesn't show this post, regardless of whether the filter is active or not ...

So, it's impossible to know whether your filter is working on this story yet, either.

tngranados 6 hours ago||||
Except it literally shows this post as the first result
lfx 6 hours ago||
I saw it after posting. Ha. That is not very smart filter, but works most of the time!
addandsubtract 5 hours ago||
Sounds like a job for AI.
ComputerGuru 6 hours ago|||
Lol, this post is number one on the leaderboard on the “filtered” list list. Trusting ai slop to filter out ai is as ironic as it gets.
hahahaa 6 hours ago|||
https://hn.algolia.com/?dateRange=last24h&page=0&prefix=fals...

or

https://lobste.rs will probably have less AI

yreg 6 hours ago||
How does one get a lobsters invite?
lfx 6 hours ago|||
You need a friend there. I'm trying to get in for years, however RO mode is still worth it.
traceroute66 5 hours ago|||
> You need a friend there.

OR you need to make a blog post that is deemed worthy.

If someone features a blog post you wrote, then you automatically qualify for access. Sort of a "right of reply".

(Features as in "new post about", not "mentioned in some thread")

deivid 5 hours ago|||
send me an email
rs_rs_rs_rs_rs 6 hours ago||||
You don't need an invite to read.
deivid 5 hours ago|||
send me an email
virtue3 6 hours ago|||
definitely take the breaks when you need them. I've already had a few friends just get lost in the AI train of stuff and suffer mentally a bit.
jmward01 6 hours ago|||
I see a future HN post about how someone vibe coded HN to filter the AI stories. HNAI (Heck No AI)
_superposition_ 6 hours ago|||
I think we have a need to revise the old let me Google that for you thing

Click the link to view conversation with Kimi AI Assistant https://www.kimi.com/share/19f6b96d-fdd2-8589-8000-0000daada...

nazgulsenpai 6 hours ago|||
Same but 100% serious
boguscoder 6 hours ago|||
Why only a half measure
More comments...