News for hackers. Stuff that matters.
Re:Kimi K3 is now live by nullbio on Thursday July 16, 2026 @12:11PM (#48936496)
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.
Re:Kimi K3 is now live by cmrdporcupine on Thursday July 16, 2026 @12:14PM (#48936533)
That depends entirely on the hosting situation. If someone can provide a subscription plan at slightly lower rates, it's absolutely compelling.
Re:Kimi K3 is now live by vidarh on Thursday July 16, 2026 @12:36PM (#48936802)
Moonshot has subscriptions maxing out at $199/month. Not home so not had a chance to see if K3 is included yet.

EDIT: Just switched my Kimi-CLI session to K3 and resumed my ongoing /goal... Will be interesting to see if I notice a difference.

Re:Kimi K3 is now live by vidarh on Thursday July 16, 2026 @05:52PM (#48940776)
I'll say after having it run for a few hours that I still don't feel it matches even Sonnet. It still does a lot of back and forth that feels dumb, but it's possible this is in effect Anthropic tricking us by hiding the full reasoning traces - who knows what Sonnet still sounds like if you were to see the whole thing.
Re:Kimi K3 is now live by cmrdporcupine on Thursday July 16, 2026 @10:00PM (#48942599)
I haven't tried K3 yet but my experience with older Kimi models was exactly this, that they'd spin for a whole chunk of time with a lot of back and forth thinking.

But some of this might still be something that gets sorted out with finding the right parameters etc on the serving side.

Re:Kimi K3 is now live by vitalyan8184 on Thursday July 16, 2026 @01:35PM (#48937634)
neither ClosedAI nor Misanthropic will let you use their models without them watching and storing the exchanges indefinitely. no sane company dealing with PII and/or trade secrets allows its employees to use those.
Re:Kimi K3 is now live by carljungslabtek on Thursday July 16, 2026 @02:13PM (#48938114)
Is this really true? I was led to believe my company had an enterprise zero data retention agreement with them and it’s why we didn’t get access to Fable

Is there proof of what you’re saying or is it just a guess?

Re:Kimi K3 is now live by vitalyan8184 on Thursday July 16, 2026 @02:37PM (#48938425)
oh, I've no doubt the US government and giga corporations can get zero data retention without ten pages of fine print. the rest of us can't.
Re:Kimi K3 is now live by lallysingh on Thursday July 16, 2026 @02:47PM (#48938580)
Unless you spend 5min googling and see that you can do zero retention via AWS Bedrock.
Re:Kimi K3 is now live by carljungslabtek on Thursday July 16, 2026 @03:58PM (#48939473)
Yeah even the chatgpt teams subscription claims ZDR. I believe the business plan from anthropic does too.

Of course maybe there is some fine print I haven’t read, and obviously I get the point that it may not be trustworthy.

edit: whoops I just checked and the “business”/“teams” plans just agree not to use your data for training

Re:Kimi K3 is now live by program_whiz on Friday July 17, 2026 @09:21AM (#48947093)
AWS Bedrock has stated that all new Anthropic and OpenAI models will not allow ZDR (only older models).
Re:Kimi K3 is now live by traceroute66 on Thursday July 16, 2026 @03:34PM (#48939190)
> zero data retention

Zero data retention is also "trust me dude".

There is no viable way of checking they are actually doing that.

That's assuming they don't put carve-out clauses in, like Anthropic did with Fable, which means data retention is back on the cards, no exceptions.

Also don't forget a zero data retention clause is still subject to the good old "law, or court or administrative order" contract clauses. :)

To get properly close to real zero-retention in a hosted model, you would have to use one of the verifiably private AI that runs in enclaves, e.g. Tinfoil (US) or Privatemode (Germany)[2]. Yes, still not the same as running on your own hardware, but a million lightyears ahead of "zero data retention" "trust me dude" clauses.

[1]https://tinfoil.sh/ [2]https://www.privatemode.ai/

Re:Kimi K3 is now live by carljungslabtek on Thursday July 16, 2026 @04:01PM (#48939503)
No I know of course, I don’t trust them as far as I can throw them when all of these companies committed the largest copyright theft in human history to build the models.

I just wanted to know if that other person had proof or not, and I guess they didn’t. I would still rather have some semblance of an agreement than not have one at all — if you’re coding on a consumer plan you should just 100% assume anything you write with it will end up in the training set

Re:Kimi K3 is now live by anon373839 on Thursday July 16, 2026 @04:39PM (#48939897)
Read the terms of the ZDR policy with a critical eye. You’ll find that Anthropic retains almost arbitrary rights to retain anything it wants.

https://code.claude.com/docs/en/zero-data-retention

Re:Kimi K3 is now live by jayzalowitz on Thursday July 16, 2026 @05:16PM (#48940357)
https://trustedrouter.com/models/moonshotai/kimi-k3 is a good option if you want that to actually be the case.
Re:Kimi K3 is now live by sheepscreek on Thursday July 16, 2026 @05:19PM (#48940395)
AFAIK there’s no ZDR with Claude models accessed directly via Anthropic. You’d have to go through either Google Vertex, Azure or AWS for true ZDR (at least legally/on paper).
Re:Kimi K3 is now live by anon373839 on Thursday July 16, 2026 @10:33PM (#48942783)
Yeah. I may be naive, but I do trust the major cloud infra providers to offer real ZDR. Though admittedly, I haven't read their terms so it's possible that they also contain egregious loopholes.
Re:Kimi K3 is now live by Arubis on Thursday July 16, 2026 @01:58PM (#48937939)
In context it seems your recommendation is to instead send those data to models within Chinese nation-network space. I’m not here to defend US frontier model companies; your accusation is probably accurate. But I doubt sending data to China is an improvement.
Re:Kimi K3 is now live by vitalyan8184 on Thursday July 16, 2026 @02:19PM (#48938191)
with open weight models, you have three other options

A) use a provider that pinky-swears not to store your data. they obviously don't give a fuck about 'distillation attacks', so they have little motivation to voluntarily monitor and store your queries. reasonably high likelihood of privacy.

B) rent the hardware and run the model yourself. very high likelihood of privacy.

C) buy the hardware and run the model yourself. absolute certainty of privacy.