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

41–50 of 1001 posts

Re: Kimi K3: Open Frontier Intelligence

#41

Post by cyanydeez »

Tiberium wrote: post #2

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…

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.

Re: Kimi K3: Open Frontier Intelligence

#42

Post by cyanydeez »

mrtksn wrote: post #19
lfx wrote: post #7

Earlier quoted context omitted.

Here you go https://tools.simonwillison.net/hacker-news-filtered

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?

Us the AI to build the bubble against the AI, because everyone knows AI is the AI of the AI.

Re: Kimi K3: Open Frontier Intelligence

#43

Post by rybthrow2 »

sixtyj wrote: post #35
Tiberium wrote: post #2

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…

[flagged]

Or the American one :)

Re: Kimi K3: Open Frontier Intelligence

#44

Post by sudosysgen »

nkmnz wrote: post #29
tw1984 wrote: post #9

> 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

> > ...ranks second only to Claude Fable 5 and GPT-5.6 Sol. So... it ranks THIRD?

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.

Re: Kimi K3: Open Frontier Intelligence

#46

Post by shrubby »

sixtyj wrote: post #35
Tiberium wrote: post #2

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…

[flagged]

Sadly these days this seems like the least worse of the three major regimes.

Re: Kimi K3: Open Frontier Intelligence

#47

Post by leecommamichael »

dghlsakjg wrote: post #23
Tiberium wrote: post #2

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…

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.

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.

Re: Kimi K3: Open Frontier Intelligence

#48

Post by freely0085 »

sixtyj wrote: post #35
Tiberium wrote: post #2

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…

[flagged]

Better than handing it over to the US regime.

Re: Kimi K3: Open Frontier Intelligence

#49

Post by petu »

Aeolun wrote: post #32
msdz wrote: post #24

> 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…

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?

It's also 2.8x parameter count (1T -> 2.8T), likely higher activation per token (50B?).

Re: Kimi K3: Open Frontier Intelligence

#50

Post by paxys »

ekojs wrote: post #37

> 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. > K…

> 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.

Post reply on HN