frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Kimi K3 is now live

https://www.kimi.com/en
286•vincent_s•1h ago•158 comments

OnePlus halts operations in USA and Europe

https://community.oneplus.com/thread/2170715118587871237
378•pilililo2•6h ago•201 comments

Microsoft Comic Chat is now open source

https://opensource.microsoft.com/blog/2026/07/16/microsoft-comic-chat-is-now-open-source/
20•jervant•22m ago•4 comments

Goes-19 weather satellite enters Safe Hold mode

https://www.spaceweather.gov/news/goes-19-safe-hold
85•yabones•2h ago•41 comments

How Our Rust-to-Zig Rewrite Is Going

https://rtfeldman.com/rust-to-zig
154•jorangreef•4h ago•54 comments

Sony Deletes a Bunch More Movies from the Accounts of People Who 'Bought' Them

https://www.techdirt.com/2026/07/15/sony-deletes-a-bunch-more-movies-from-the-accounts-of-people-...
293•nekusar•4h ago•159 comments

Let's Build PlanetScale from Scratch: Infrastructure

https://onatm.dev/2026/07/16/homescale-part-1/
85•onatm•4h ago•16 comments

GC shape stenciling in Go generics

https://rednafi.com/go/gc-shape-stenciling/
11•ingve•4d ago•0 comments

Guide to data tools landscape for developers

https://sinja.io/blog/data-landscape-guide-for-developers
26•OlegWock•1h ago•2 comments

Ente – Opening Our Books

https://ente.com/open/
143•Sherex•5h ago•39 comments

Where are YC founders now? OpenAI and Anthropic, mostly

https://joinedanthropic.com
266•ohong•8h ago•162 comments

Show HN: I've built a words game based on binary search

https://hilogame.cc/
37•ludovicianul•3h ago•38 comments

The lost joy of music piracy

https://www.pigeonsandplanes.com/read/music-piracy-what-cd-oink-nine-inch-nails-streaming
610•mcgin•11h ago•406 comments

Inkling: Our Open-Weights Model

https://thinkingmachines.ai/news/introducing-inkling/
1141•vimarsh6739•22h ago•278 comments

Accidental Anonymity

https://macwright.com/2026/06/24/accidental-anonymity
20•caminanteblanco•2d ago•2 comments

How to Train a Gen AI Kick Drum Model on Your Old Linux Desktop with 6GB VRAM

https://www.zhinit.dev/blog/training-a-kick-drum-diffusion-model
11•zhinit•1h ago•5 comments

Teardown: A Generic 7-Port USB 3.0 Hub That Wasn't

https://goughlui.com/2026/07/09/teardown-a-generic-7-port-usb-3-0-hub-that-wasnt/
168•speckx•3d ago•73 comments

1,300 Beautiful Wildlife Illustrations from the 19th Century Now Restored

https://www.openculture.com/2026/07/explore-1300-beautiful-wildlife-illustrations-from-the-19th-c...
190•gslin•13h ago•31 comments

Let's build a simple interpreter for APL – part 1

https://mathspp.com/blog/lsbasi-apl-part1
32•mpweiher•6d ago•0 comments

Grok Build is open source

https://github.com/xai-org/grok-build
555•skp1995•20h ago•585 comments

The Act and the Outcome of Creation

https://www.ssp.sh/blog/on-creation/
23•zazuke•4h ago•5 comments

Governments, companies, nonprofits should invest in free, open source AI [pdf]

https://www.siegelendowment.org/wp-content/uploads/2026/07/fortune-david-siegel-open-source-ai.pdf
268•bilsbie•19h ago•95 comments

Track your workout from the iPhone Lock Screen

https://musklr.com/blog/2026/iphone-lock-screen-workout-tracking-live-activity/
31•badgag•5h ago•28 comments

Panel meter calculator with floating point

https://lcamtuf.substack.com/p/panel-meter-calculator-with-floating
28•surprisetalk•3d ago•5 comments

My Homepage Has a Pulse

https://snehankekre.com/posts/hr/
7•punk_ihaq•2h ago•0 comments

If you want to create a button from scratch, you must first create the universe

https://madcampos.dev/blog/2026/07/accessibility-from-scratch/
207•treve•12h ago•100 comments

Introduction to KizunaShelf: A shelf for everything you love

https://mudkip.me/2026/07/16/Introduction-to-KizunaShelf/
21•mudkipme•3h ago•4 comments

Reynard: A real Firefox web browser for iOS 13 or later

https://github.com/minh-ton/reynard-browser
131•AbuAssar•11h ago•35 comments

World-War-Ⅱ-era telephone line still in use in Upper Tanana Valley Alaska (2021)

https://www.sketchesofalaska.com/2021/03/world-war-ii-era-telephone-line-still.html
19•Lammy•5d ago•3 comments

SQLite should have (Rust-style) editions

https://mort.coffee/home/sqlite-editions/
338•gnyeki•17h ago•154 comments
Open in hackernews

Kimi K3 is now live

https://www.kimi.com/en
275•vincent_s•1h ago

Comments

Tiberium•1h ago
More details:

- https://platform.kimi.ai/docs/guide/kimi-k3-quickstart

- https://platform.kimi.ai/docs/pricing/chat-k3

1M context, pricing is $3/$15 for 1M tokens (cache $0.3), which is extremely high for a Chinese open-weight model, but if it's truly competitive with most of the current frontier and is only behind Fable/Sol, the pricing is justified.

This is 1:1 pricing of Anthropic's Sonnet series (except Sonnet 5 which is currently on discount), and very close to 5.6 Terra pricing (Terra's input is $2.5).

One thing to consider, though: reasoning efficiency matters directly for how expensive a model actually is in real use. GPT's models are extremely reasoning efficient, and some Claude models like Fable at lower effort are as well. So if Sol spends 10K reasoning tokens to do something (at $30/1M) vs Kimi K3 that spends 50K reasoning tokens, Sol would win on cost effectiveness.

Deukhoofd•1h ago
I feel like the quickstart is missing something. It's referring to its tech blog for actual benchmarks, but K3 isn't mentioned on there, the last thing on that blog was K2.6, 2 releases ago.
dghlsakjg•1h ago
Tokenizers also matter. Anthropics tokenizers will encode the same piece of text at a way higher token count than OpenAi, for example.

That said, Kimi is competing against GLM in my mind, and GLM 5.2 is less than 1/3 the price.

asenna•1h ago
With that kind of pricing, I don't think they're competing with GLM with this new launch.
leecommamichael•1h ago
Tokenizers define the alphabet on which the language model is trained. I don't want people to get the impression it's a module which can be swapped out or modified on its own. Alphabet size is a design consideration related to correctly encoding the training data.
smallerize•25m ago
That's true, but it makes it difficult to compare pricing when it's based on tokens. Maybe we need a benchmark for price per a specific input, like enwiki8.
leecommamichael•5m ago
Yes, almost all work people share which seeks to measure the capabilities and differences of models needs to get more precise. We are clamoring to say something meaningful about these things.
cmrdporcupine•13m ago
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 work because the monthly coding plans from Anthropic and OpenAI are a better deal.

martinald•1h ago
Will be interesting to see how it stacks up pricing wise on the various inference providers.
cyanydeez•1h ago
I eat 1M context in a local model in about 3-4 hours.

It'd need to be exceptionally smart and error free to ever make sense.

mmaunder•53m ago
Agreed re reasoning. I’ve seen this play out with 5x reasoning negating cost savings.
schmorptron•48m ago
Are thinking models only the reasonable tradeoff vs using much larger non thinking ones because the cost of output tokens is below that of input tokens?
csomar•46m ago
It seems the subsidized era is nearing its end and we'll see a convergence on API pricing before a pulling of subscriptions pricing.
easygenes•28m ago
That’s not what this indicates. This is the biggest and most expensive to serve, and most capable open weights model yet. They’re just pricing it in line with capabilities.

Kimi also offers generous subscriptions. Subs aren’t going anywhere. Think of subs like running an insurance business. There might be some users you lose money on (ones who max out their weekly quota without fail), but they’re managed such that the average subscription turns a healthy profit. There’s never been subsidies in model serving, inference is just cheaper in terms of ops TCO than people assume, and API margins are very high.

nullbio•17m ago
Ah, the old "subsidized" meme always rearing its head. Yawn.
0xbadcafebee•32m ago
The big danger here is the gradual increase in open-weight subscription costs. I use open weight subscriptions, with lower-cost models for 80% of my tasks and GLM-5.2, Qwen 3.7-Max, Kimi-K2.6/2.7-Code for the 20% that need the most intelligence. That lets me maximize the rate-limit the subscription gives (rate limits per model are literally a price-limit-per-token/model). When new/more expensive open weights come in, providers phase out older/cheaper models. Over time we will either have to pay more, or use our subscriptions less.

It goes without saying, but if the open weights become as expensive as SOTA models, there's no point in using open weights. If nobody pays for open weights' development, the development dies out, and we're stuck with a US-controlled duopoly again. Which may be the biggest threat the world has seen from the US since nukes.

h14h•21m ago
> reasoning efficiency matters directly for how expensive a model actually is in real use

I have high hopes on this topic, given token efficiency seemed to be the primary (only?) goal of the K2.7 Code release.

Excited to see the signals that come out of the big eval/benchmark sites.

nullbio•16m ago
This is too expensive to be a viable model. If it were $5/1m output, it might be another story. At these prices, there's no reason to use this over GPT 5.6.
cmrdporcupine•14m ago
That depends entirely on the hosting situation. If someone can provide a subscription plan at slightly lower rates, it's absolutely compelling.
sroerick•16m ago
How do Kimi's subscriptions work? I find their price structure pretty confusing
esher•1h ago
Half kidding feature request for HN: Mark all AI related posts so I can filter them out, when I need a pause.
lfx•1h ago
Here you go https://tools.simonwillison.net/hacker-news-filtered
tngranados•1h ago
Except it literally shows this post as the first result
lfx•32m ago
I saw it after posting. Ha. That is not very smart filter, but works most of the time!
addandsubtract•4m ago
Sounds like a job for AI.
mrtksn•1h 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•1h ago
tw1984•1h ago
> Among the models tested, its overall intelligence ranks second only to Claude Fable 5 and GPT-5.6 Sol.

> The full model weights of Kimi K3 will be released in the coming days. More details on the architecture, training, and evaluation will be published together with the Kimi K3 technical report.

https://platform.kimi.ai/docs/guide/kimi-k3-quickstart

nkmnz•1h ago
> > ...ranks second only to Claude Fable 5 and GPT-5.6 Sol.

So... it ranks THIRD?

polski-g•1h ago
USSR is proud to announce that they won 2nd place in an Olympic contest. The filthy USA regime? Next to last!

(There were only two countries competing in said event)

amelius•54m ago
Apple proudly announced they won 2nd place in a competition among smartphone OSes.
yreg•41m ago
Apple would never claim to be second.
ayushpai•16m ago
1st in open weight
blovescoffee•1h 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.
kamranjon•1h ago
Where did you hear about the deepseek release? Would love to follow the same source.
blovescoffee•23m ago
They emailed current paying users of the api (or at least that’s how I got updated).
bayesianbot•1h 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.

hack1312•56m ago
What provider are you using?
bayesianbot•54m ago
DeepSeek's own API
greyb•
khalic•1h ago
I really need to finish my automated model evaluation harness, I can't keep up with this pace
GodelNumbering•1h ago
I've playing around in between with Arc-AGI-3 lately. Based on my very quick test prompt, I do not think it will achieve any meaningful score in Arc AGI 3. Not that it was expected to.
msdz•1h 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?

Aeolun•1h 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?
petu•58m ago
It's also 2.8x parameter count (1T -> 2.8T), likely higher activation per token (50B?).
pixl97•43m 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.

buildbot•1h 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)

NoImmatureAdHom•44m ago
Surely it's only open weights?
stefan_•15m ago
It's not even that right now.
kroaton•16m ago
It also goes to show that Fable/Sol must be 4-5T in size.
wxw•1h ago
Open source Fable/Sol challenger! Interesting to do a release product-first.

https://platform.kimi.ai/docs/guide/kimi-k3-quickstart

ekojs•1h ago
> In our evaluations, Kimi K3 delivers frontier-level performance. Among the models tested, its overall intelligence ranks second only to Claude Fable 5 and GPT-5.6 Sol. For the complete benchmark results, see our tech blog. The full model weights of Kimi K3 will be released in the coming days. More details on the architecture, training, and evaluation will be published together with the Kimi K3 technical report.

> K3 pushes the boundary of end-to-end knowledge work. On the GDPval-AA v2 leaderboard, Kimi K3 scores 1687. The benchmark evaluates AI models on real-world tasks across 44 occupations and 9 major industries; Kimi K3 ranks behind only Claude Fable 5 Max and GPT-5.6 Sol Max, and ahead of Claude Opus 4.8 Max at 1600.

> On AA-Briefcase, Kimi K3 scores 1527, ranking second among all models — behind only Claude Fable 5 Max and ahead of GPT-5.6 Sol Max (1495). AA-Briefcase is a private agentic knowledge-work benchmark developed by Artificial Analysis to evaluate frontier agentic capability in long-horizon knowledge work.

Really good benchmark score it seems. Maybe another DeepSeek moment right here.

paxys•58m ago
> its overall intelligence ranks second only to Claude Fable 5 and GPT-5.6 Sol

Pretty sure ranking “second” to two others means ranking third.

ekojs•52m ago
Yeah, bad wording it seems. Though a charitable interpretation is that Fable 5 and GPT 5.6 Sol are joint 1st place in the measurement.
paxys•49m ago
Doesn’t matter, the next one is still third.
cheesecakegood•
smalltorch•1h ago
Account creation with only a phone number or google account is lame.
kleiba2•52m ago
Especially if you don't have a phone and don't want to use your google account for anything but gmail, for privacy reasons. Both of these point apply to me, for instance.
lvl155•1h ago
Say what you want about these Chinese models but they sure create competition and urgency in the space.
_superposition_•43m ago
Agreed, this will save us all money in the long run.
antiloper•56m ago
Seems to only use ≈60% as many reasoning tokens as 2.6. So the price hike is not as bad as it looks.
schmorptron•54m ago
That's a more than 2x jump in parameter count. I know it's not a measure of quality by itself, but it will be interesting how it "scales". Bust it looks like they're gonna be competing with the big boys now, pricing also approaches Gpt 5.6 Terra
pr337h4m•52m ago
It does seem to have retained the K2 series's creative writing abilities, at least with the prompts I've tested so far.
satvikpendem•52m ago
Now, will they actually release the weights? Seems like Chinese model providers are slowly closing up, like Alibaba's Qwen 3.6 which did release weights (but not the biggest parameter count ones) and none for 3.7.
j2j8•27m ago
In the coming days
xyzsparetimexyz•51m ago
Any updated Pareto frontier graphs? https://paraplouis.github.io/llm-pareto-frontier/ is quite out of date now.
tao_oat•29m 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.

Bromeo•28m ago
openrouter->rankings shows a pareto frontier. https://openrouter.ai/rankings#benchmarks
tskj•47m ago
I'm curious if they're keeping up mostly due to distillation or how that works. Does anyone outside China know?
XCSme•46m ago
No blog post? Benchmarks?
dmix•44m ago
This might have been published before they released their tech blog, I don't see one
naaqq•31m ago
Will be later.
npn•45m ago
Not worth it. I have just tried a single prompt in the web interface and it is still not finish reasoning. It thinks too much and often repeats the same stuff over and over.

Combine with the price it will surely more costly than gpt 5.6.

verdverm•5m ago
Its bad to judge these things on immediate release, there is a spike of excited users and that distorts performance. Also bad to judge from on a single interaction, you'll get bad requests with every provider, super busy times raise the probability
m3h•44m ago
> Kimi K3 is Kimi’s most capable model to date, with 2.8 trillion parameters.

This puts them on the top of the largest open models list:

  Kimi K3            2.8T
  DeepSeek-V4-Pro    1.6T (49B active)
  Kimi K2.6          ~1T (32B active)
  GLM-5.2            754B (40B active)
  DeepSeek-V3.2      685B
  Mistral Large 3    675B
That's one mighty large model! Moonshot is going to need the USD 500 million reportedly raised earlier this year to run this model.
wolttam•40m ago
I guess it remains to be seen whether this will be open-weights. We don't even know how many active params at this point.
sudosysgen•37m ago
The article says weights will be released in the coming days, and hints it's likely around 50-70B active params.
wolttam•33m ago
It did say that, but it doesn't any longer.
simonw•20m ago
What's the URL of the article that used to say that?
wolttam•15m ago
https://platform.kimi.ai/docs/guide/kimi-k3-quickstart this one, it used to have more information about the model itself, similar to the K2.6 and K2.7 pages.

Edit: OpenRouter still describes it as an open-weight model: https://openrouter.ai/moonshotai/kimi-k3

Guess we'll see!

ncruces•39m ago
I get a quota of GitHub Copilot for free.

From all the models available to me I'm most happy with Kimi K2.7 (given the cost/performance).

wolttam•33m ago
I'm a bit nervous this one isn't going to be open-weights. Any mention of "open" has been struck from the literature for this model (it was present an hour ago). We don't even know active params?

At this pricing, I'll be surprised if it's open.

nullbio•14m ago
This does seem like a cash grab. These token rates are crazy. I'll just use GPT 5.6 thanks.
icedrift•12m ago
Reuters has been reporting that Chinese government is undergoing similar investigation to the US; blocking the export of domestic frontier models. They boil down to "anonymous sources" but it does seem inevitable as the tech gets stronger and stronger.
simonw•31m ago
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.

eleventen•27m ago
Oof, front fork is wrecked. Pelican should be wearing a helmet on that death trap.
simonw•26m ago
I like that it has a snazzy red scarf.
ryanseys•16m ago
I appreciate the tiny flowers in the grass.
bitexploder•25m ago
It is a nice pelican, though. At least it has that going for it.
smallerize
XCSme•29m ago
Only supporting "max" reasoning is weird, their parameters are quite inflexible atm:

    Important limits:

    reasoning_effort currently supports only max; K3 always has thinking mode enabled.

    max_completion_tokens defaults to 131072 and can be set up to 1048576.

    temperature=1.0, top_p=0.95, n=1, presence_penalty=0, and frequency_penalty=0 are fixed; omit them from requests.

    Return the complete assistant message unchanged in multi-turn conversations and tool calls.

    Vision input does not support public image URLs. Use base64 or ms://<file-id>, and make content an array of objects.

    Web search is being updated and is not recommended for production workflows in the near term.
anthonypasq•26m ago
Does anyone have any heuristics on how scaling parameter count actually scales cost to serve? Also im assuming we dont really know the sparsity here?

Is them pricing at Sonnet level actually give us any information at all at how big Sonnet is or is there too much opacity around inference margins?

nullbio•19m ago
This is far too expensive. Why would I use this over a frontier model at these prices.
cute_boi•17m ago
Thank you Kimi. We no longer need to rely that much on Dario and his supreme lackeys to decide what is safe or not for simple tasks.
Gecko4072•13m ago
Kimi K2.5 and K2.6 have always missed that extra piece of magic that I instantly noticed even in GLM 5.1. Strange how these things you can’t quantize make such a big difference. Curious what other people’s impressions are.
XCSme•12m ago
I am trying to benchmark it, but it only supports (max) reasoning, and even for simple questions, it takes forever to answer/times out :(
Us the AI to build the bubble against the AI, because everyone knows AI is the AI of the AI.
postalcoder•1h 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•54m 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.

ComputerGuru•49m 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.
boguscoder•1h ago
Why only a half measure
nazgulsenpai•1h ago
Same but 100% serious
jmward01•1h ago
I see a future HN post about how someone vibe coded HN to filter the AI stories. HNAI (Heck No AI)
hahahaa•1h ago
https://hn.algolia.com/?dateRange=last24h&page=0&prefix=fals...

or

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

yreg•43m ago
How does one get a lobsters invite?
lfx•21m ago
You need a friend there. I'm trying to get in for years, however RO mode is still worth it.
rs_rs_rs_rs_rs•21m ago
You don't need an invite to read.
virtue3•57m 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.
_superposition_•46m 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...

sudosysgen•1h ago
The literal interpretation of that sentence is "when it is second or third, it is only behind Fable 5 or 5.6 Sol". And indeed they give benchmarks where it is ahead of one but not both models.
markasoftware•26m ago
They've removed the paragraph about releasing model weights.
xur17•22m ago
Does that mean this one won't be open source?
45m ago
Any way to avoid China sales tax or is that just the cost of doing business?
NortySpock•40m ago
https://openrouter.ai/deepseek/deepseek-v4-pro#providers

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

verdverm•8m ago
Fireworks.ai
sudosysgen•48m 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•22m 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?
petu•6m ago
Deepseek V4 paper is just ~three months old
surgical_fire•8m ago
Ohh I didn't know about it. Finally something to be excited about.
7734128•48m 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•28m ago
TIL, that makes a lot of sense, and thanks for the correction.
4m ago
DENSE_RANK() vs RANK() claims another victim
jnwatson•48m ago
If there are two folks standing at gold, nobody gets the silver medal.
worldthruword•23m ago
But linearizing an equal magnitude quantities by alphabet priority would be unfair. Magnitude is the important quantity here.
scotty79•51m ago
Which is still great because it means neither of the two best financed labs in the world manage to produce even two models themselves that would beat Kimi K3.
laybak•26m ago
good to call out this sleight of hand in their wording
akoumjian•54m ago
Where are you seeing this write up?
ekojs•49m ago
I copied that from https://platform.kimi.ai/docs/guide/kimi-k3-quickstart but it seems they updated the page to remove the benchmark score now.
Aurornis•48m ago
> > K3 pushes the boundary of end-to-end knowledge work. On the GDPval-AA v2 leaderboard, Kimi K3 scores 1687. The benchmark evaluates AI models on real-world tasks across 44 occupations and 9 major industries; Kimi K3 ranks behind only Claude Fable 5 Max and GPT-5.6 Sol Max, and ahead of Claude Opus 4.8 Max at 1600.

This is the same benchmark where Sonnet 5 outperforms Opus 4.8 max.

Like all model releases, the benchmarks aren't going to tell the whole story. All of the open weight models come with amazing benchmark results now. It's hard to believe anything other than that the benchmarks are leaking into (or intentionally included) into training data.

rd•36m ago
i’ll never really understand this comment. why would labs do this if they know private benchmark evals will come out in the next week?
kroaton•17m ago
Ling/Ring 1T-A50B and the new Inkling 975B-A41B deserve to be on that list.
•
21m ago
How did "Generate an SVG of a pelican riding a bicycle" turn into 95 tokens?
simonw•19m ago
That's a great question.

I just tried "hi" through the same OpenRouter API and the input token count for that was 86 - and for "hi there" the count was 87.

I think there's an 85 token hidden system prompt of some sort.