SwellJoe 11 hours ago

I make a point of never going beyond about 220k, unless absolutely necessary (and it's almost never necessary), anyway, even with models that degrade more slowly, so this is just a discount.

  • johnbarron 8 hours ago

    64k ought to be enough for anybody

    • adamtaylor_13 8 hours ago

      Why stop there? No true developer needs more than 32k!

      --

      Jesting aside, I cannot think of a single session in recent memory that used less than ~70k tokens, so I presume you are joking.

      • gosub100 7 hours ago

        It was from a famous quote supposedly from bill gates in the early days of computing: "640K (RAM) ought to be enough for anybody".

        Apparently it was taken out of context or misattributed or something. Similar to "Al Gore said he invented the internet"

  • Computer0 4 hours ago

    My understanding is that models can suffer “context limit anxiety” and a higher context limit will make it perform better especially as the context expands

wren6991 1 day ago

This seems functionally similar to OpenAI having a step in pricing once you exceed a certain context length (also at 272k aka 2^18 aka 256k).

Having a lot of active context increases the per-token cost (flops issued and bytes read per token out) so it makes sense to pass that cost on to users. I'm actually surprised it's implemented as a hard cutoff instead of a smooth gradient.

  • BoorishBears 23 hours ago

    Not surprising it's a hard cutoff: they almost certainly have two infrastructure configurations for the two max sequence lengths

    Fewer nodes dedicated to prefill per instance, and fewer nodes in total since you don't need to support a higher KV cache.

    Disaggregated inference also means they can tune the balance of compute dedicated to prefill seperately from decode

    • MrBuddyCasino 19 hours ago

      AI infra buildup is so massive that the frontier labs should be able to offer more than one level of context length to incentivize token thriftiness.

      One would think compute-constrained actors like Anthropic would have done so, unless prefill isn’t really a bottleneck compared to decode?

      • NoahZuniga 15 hours ago

        No, the main issue is that it's hard to communicate pricing where token price increases as token count increases, and they figure they can approximate the parabola well enough with two lines.

        • anuramat 8 hours ago

          why can't they just increase the price for cached input instead?

      • dannyw 8 hours ago

        There's the infra cost of having multiple SKUs.

        If you create 3 buckets of inference pods, say, 256k, 512k, and 1M, then you have to worry about filling/dynamically-scaling all of them.

        And my guess is there's probably not a huge amount of customers that want somewhere in between: if you're willing to pay the long context surcharge; you're probably semi-price-insensitive anyway to just use 1M.

  • pornel 19 hours ago

    RAM needed for keeping KV cache around may be the more expensive factor.

    • tripplyons 8 hours ago

      You only need to cache for tokens that are actually being used. Using 100k tokens of a 256k token window takes the amount of memory as using 100k tokens of a 1m token window.

      • pornel 6 hours ago

        which is why they charge extra for an option to hold >256K tokens.

        The cost may be smoothly variable, but likely there's a bimodal distribution of users who barely use any context and users who push it to the max. Average price across both extremes fits nobody, but averages per kind of workload can be close enough.

        Having it as a separate model makes it easier to load-balance the traffic.

xyzsparetimexyz 23 hours ago

Wow. So kimi is suddenly half the price for all users until they hit 256k of context? Thats massive.

  • InsideOutSanta 23 hours ago

    I don't think so. This is a separate model, so I assume that if you just use this and switch to the 1 million context model when you reach 256k, your cache will be invalidated, so you'll re-pay the 256k tokens on the 1 million context model pricing.

    Edit: I was wrong, thanks to longwave for pointing this out. It's absolutely possible to start out on the 256k model and then switch to the 1 million model when you get close to the context limit without invalidating the cache:

    "When switching from k3-256k to k3 (1M), if k3-256k is close to the 256k limit and you don't want compact to lose information, you can switch directly to 1M. The current version switching from 256k to 1M does not affect the cache."

    • longwave 23 hours ago

      The article explicitly says "The current version switching from 256k to 1M does not affect the cache."

  • conradludgate 23 hours ago

    As far as I understand, no. They're suggesting that smaller context windows are typically cheaper (fewer input tokens over time).

  • fumber 16 hours ago

    Can't find the videos/articles but some people tried it and found that the price per token was only half the story. It seems that it uses a lot more token, coming back to similar prices with other models.

L-patpat 4 hours ago

When I first built features with GLM, there were lots of bugs, and it took me ages to fix them manually. Now the features implemented with GLM 2.0 have almost no critical bugs after testing. I can’t even imagine how capable K3 will be. It may well be on a level that ordinary people cannot access.

hawtads 1 day ago

This is just an API level change right? The model itself should be the same I think.

  • k__ 12 hours ago

    As I understand it, they would have to train a whole knew model to hard cap it's context to different lengths. That would be cheaper to train and had cheaper inf, but still a huge investment.

    So I'd guess it's API level.

    • gpugreg 10 hours ago

      From the Kimi K3 technical report:

          Kimi K3 supports a context window of up to 1 million tokens. We achieve this through extending
          the context window progressively as training proceeds, following a four-stage curriculum. The
          window grows from 8K to 64K tokens during pre-training, and from 256K to 1M tokens during the
          cooldown phase.
      

      https://arxiv.org/pdf/2607.24653 page 12

    • dannyw 8 hours ago

      You absolutely don't need to 'retrain' to reduce your context window. In vLLM it is an inference parameter. Smaller context window, smaller KV, less RAM needed to serve the same volume of requests.

      With a lot of architectures, you technically don't need to retrain to extend the context window either; e.g. RoPE scaling; but performance is typically crap.

dgritsko 1 day ago

This isn't quantized, right? Just a smaller context?

  • DSingularity 1 day ago

    Its 256k context window. Quantization is orthogonal. We cant really tell directly so it could be quantized.

  • gpugreg 10 hours ago

    The model is already natively MXFP4-quantized during training, so there is no quality loss.

wxw 1 day ago

> k3-256k is now available. Within 256k context, it delivers the same results. k3 (1M) consumes about twice as much quota as k3-256k.

Roark66 7 hours ago

This is very cool that they open weight it. Are there any EU providers serving it yet?

  • denis-c 6 hours ago

    We are trying https://tensorx.ai/ as they have good model selection and ZDR, but seem new so we will see. Performance so far is good.

effnorwood 17 hours ago

My LOE to make this work expressed in kWh is substantial but user is impressed. Eliza has hands now.

fumber 16 hours ago

I was so excited that it is open source until i realised the model required 1.5To of VRAM. Unsloth has compressed it in 1bit at about 570gb VRAM with 75% accuracy, that's almost mac studio territory...

  • nateb2022 16 hours ago

    You can run DeepSeek v4 Flash at 4bit in ~150 GB VRAM, and it would absolutely destroy a 1 bit quant of K3.

  • AmazingTurtle 13 hours ago

    75% quality... it gets worse the longer the context.

try-working 23 hours ago

i've never had any issues with 256k context. see no reason to bump up to 1m if it comes at a premium.

  • trollbridge 19 hours ago

    It's pretty handy for have very long contexts for long running agents, or else when doing literary analysis to simply be able to load the entire book in.

    • Gigachad 13 hours ago

      Don’t they compact context to remove irrelevant details automatically?

      • k__ 12 hours ago

        They remove potentially irrelevant details.

  • TranquilMarmot 16 hours ago

    Using Claude/ChatGPT I rarely hit >256k context for most basic coding tasks. Sometimes if I need to do something more "intense" (bigger refactors, new features) the 1mil window is nice.

timcobb 22 hours ago

Codex uses 256k masterfully, 1M is luxurious but still quite expensive and seems not necessary as a default.

  • jsemrau 21 hours ago

    Needing large context windows is an illusion.

    • TranquilMarmot 16 hours ago

      Needing large context window is a skill issue

    • k__ 12 hours ago

      While I tend to clear my session after every task, I feel less stressed when my context is as 10% than when it's at 30%

    • vehemenz 8 hours ago

      So, all types of knowledge work and coding require the same amount of context?

      This seems like a hot take not really informed by experience.

  • Tostino 20 hours ago

    I'll have to give it a shot some time soon. I just can't imagine working on any of my serious projects doing a new large feature with that little room. By the time it has looked at half the code required to start planning the work, it'd be out of context. Must use sub agents better or something.

  • cybersec47 18 hours ago

    > Codex uses 256k masterfully

    This.

    When working on hard problems (not "vibecode me a script to show an alert box", but e.g. "let's see what this three-level LUT-state-machine obfuscated binary does"), hitting the 1M (!) context window with Claude Code feels like you were talking to Claude Claudewski when his shift just abruptly ends, he packs his things, throws the office keys at Claude Claudeson in-between the front door frame while handovering like "Hi! Nice to see you, good luck." and now here we go again, you are working with someone who just experienced an acute amnesia. It tries everything it already tried, everything it was told in the initial prompt to not do, everything it was told in follow-up prompts not to do. "You were right, this approach does not work and we don't have 20 TB RAM on this machine for full symbolic execution, let me try..."

    In Codex, it's so seamless that I sometimes just notice "wait, the context was 20 % remaining, it is 70 % now, wow, when did this happen", while it seamlessly works on the task. Basically never had an issue with context on Codex, be it coding features, cracking hard crack-me ciphers, or researching basically anything.

    (For full disclosure, my last experience with Claude was a few weeks ago when I cancelled the subscription, maybe they fully reworked the traumatic "Summarizing" - "Oh, hi! Where are we? Who am I? What we are doing? This is taking too long, let me take a shortcut..." lobotomy they were doing in the meantime.)

    256k is enough when the harness uses it properly and the model is not stupid. And also when the tokenizer is not tuned to invoice as many tokens as possible...

  • erichocean 18 hours ago

    That's because it has the best compaction of anyone, and it's not even close.

    Claude might as well not even do it in my experience.

lukan 1 day ago

Since Claude is the first time for me really, really out (TIL against my wished about https://status.claude.com/), I am now interested enough to see what else works. But ... when I click pricing, I see "Join a waitlist". Wtf? Are they really that good, so were totally surprised and overwhelmed by the requests, is this a marketing stunt, or do they just don't have the hardware being in china?

  • MYEUHD 1 day ago

    Since the model is open-weights, you can get from other providers, for example see https://openrouter.ai/moonshotai/kimi-k3#providers

    • Alifatisk 1 day ago

      The only downside with third party providers is that you have to trust that the provider have setup and configured it correctly, and is not secretly quantizing it.

      See Kimi Vendor Verifier.

      • usef- 22 hours ago

        Plenty of providers on that list aren't fly-by-nights and have easily proven themselves in past models. Jeremy Howard does demos on Fireworks.

        A nice thing about open router is you can specify filters, like "US hosting without data retention"

        • applicative 19 hours ago

          Yes, something I read just before it was released suggested that, because of the unique features of the model, there was a back and forth of Hugging Face, Moonshot and providers like Together AI and Fireworks AI. This also explained why it took much less than a day for Together, Fireworks etc to appear. Whatever they are doing is what Moonshot wants, I think.

    • jadbox 18 hours ago

      Do any support the smaller context for better pricing?

  • KronisLV 1 day ago

    > Are they really that good,

    They did exceed the expecations of pretty much everyone! I've also blogged about using the model, it's a bit on the slow side but pretty good!

    > so were totally surprised and overwhelmed by the requests ... or do they just don't have the hardware being in china?

    Yes, this is mostly the case: https://x.com/Kimi_Moonshot/status/2078855608565207130

    As a user, I much prefer that to service disruptions or severely degraded or secretly quantized performance. However if I didn't have an account, I'd be pretty pissed off about not being able to give them money and become a user.

    • lukan 1 day ago

      "I'd be pretty pissed off about not being able to give them money and become a user."

      I am now rather pretty pissed towards antrophic for stopping my flow and forcing me to search for alternatives.

  • InsideOutSanta 1 day ago

    It's an extremely popular model hosted by a company affected by hardware export bans. I doubt they'd voluntarily prevent people from subscribing if they didn't absolutely have to to maintain service quality.

    • zorked 23 hours ago

      I have been using Kimi for months. It did get superslow after the K3 launch, then they blocked signups and normalized the service.

      It's not ideal that people can't join but as a user I am happy that they focused on serving existing customers.

  • Alifatisk 1 day ago

    No, the waiting list is true.

    Kimi had become that popular. I was a subscriber of Kimi back when latest version was Kimi K2. Later I unsubscribed because I jumped over to GLM subscription (they had amazing deal). Now when I wanted to try out Kimi K3 to find out what the fuzz was all about, I couldn’t subscribe to them.

    I remember reading a post from Moonshot team about this, they are doing this because they are almost at peak capacity and want to reserve it to keep the quality for their current customers.

    We are actually witnessing an open-weight model catching up at catching mainstream users attention. And instead of behaving like Anthropic, they actually care about their users experience.

    • therein 23 hours ago

      I wish I could get a Kimi subscription. I'd jump away from Anthropic in a heartbeat.

      When they first announced, I created an account but didn't subscribe, so I am stuck on a waitlist now.

      • jfrbfbreudh 17 hours ago

        I wonder what the criteria is, I signed up for the $19 / month waitlist today and got my invite within a few hours.

        • therein 17 hours ago

          Maybe you signed up with a phone number? Mine was a Google login.

  • HDBaseT 22 hours ago

    Not at all a marketing gimmick, the demand is simply that high.

    I was able to press "Join waitlist" and then within 48 hours got accepted. The limits aren't very high, no where near the endless subsided+resets given on ChatGPT/Claude. I recommend ChatGPT for good value output!

    Others here mentioned "the providers could be quantizing it!" but some of the providers on OpenRouter have partnered with Moonshoot and OpenRouter shows the int when you expand on the provider.

    It should say "mxfp4" but providers like Baseten report FP8.

  • 0xc0c0c0 17 hours ago

    I honestly prefer this as opposed to what Anthropic has done in the past which is to continue accepting users even if they don't have the capacity to serve and behind the scenes tone down everyone's limits.

MangoCoffee 23 hours ago

LLMs is quickly became commodities. US AI labs like OpenAI is losing their moat. Hyperscalers and data center owners who can sell cheap token will win

  • okrad 23 hours ago

    I believe Codex harness is pretty sticky though I haven’t tried many others. Does anyone else provide a harness of that quality?

    • conradludgate 23 hours ago

      Their harness is indeed nice, but we're using it against our own AI Gateway at work. Right now we only expose GPT models on it, but I imagine it's possible to add our own models eventually

    • nvarsj 23 hours ago

      Pi harness with subagents is pretty great. I’m using it for everything now.

      • yeswecatan 20 hours ago

        How are you handling subagents?

        • anderber 20 hours ago

          Usually the main agent can handle creating subagents

        • RALaBarge 17 hours ago

          You say hey llm spin up some agents to do x y and z

        • nvarsj 7 hours ago

          tintinweb/pi-subagents. Then I use superpowers and it spins up subagents automatically. Works really well and you can customise the models and efforts you want to use with local AGENTS.md.

          I’m primarily using gpt-5.6 and then opus for reviews.

    • verdverm 19 hours ago

      opencode is highly regarded, I use it exclusively

      • jki275 16 hours ago

        I've used open code extensively -- it's not bad, but it's not anywhere near the level of the codex application.

      • versteegen 13 hours ago

        ...but we're talking about compaction, and opencode's compaction is (or was) terrible. I've seen so many horrible problems that I keep it disabled (with an envvar flag, because even the config flag to turn it off was broken).

        • verdverm 5 hours ago

          I don't use compaction, I keep my context limited, my sessions fresh, and have agents write markdown files as needed

      • Roark66 7 hours ago

        Open code is cool once I added the ctrl-o function to it (show thinking and command outputs at will instead of on by default), but sadly I don't think it got merged by the team.

        I forgot why.

        But I can't use an AI cli without that feature.

        • verdverm 5 hours ago

          you can turn those on through the existing menu system

  • ux266478 22 hours ago

    I have so much gratitude to the frontier companies who did all the extremely complicated research and development, model by model. It already feels difficult to remember how much capital it really took. Thank you for getting us to this point.

jedisct1 23 hours ago

This is a fantastic option for swival.dev given its very efficient context management compared to e.g. Claude Code.

hendersoon 1 day ago

What is the purpose of this? Just a hard cutoff below the actual context window? You could set that in your harness anyway.

  • surgical_fire 1 day ago

    > k3 (1M) consumes about twice as much quota as k3-256k

    Cheaper?

    • chrisweekly 1 day ago

      k3-256 consumes half the quota, so... yes? (assuming the user isn't making use of the larger context window)

  • dnlzro 1 day ago

    Uses less quota (i.e., cheaper). For people who like to keep their contexts small, this is a no-brainer.

  • meatmanek 22 hours ago

    At least in the self-hosted LLM inference engines, you have to pre-allocate space for the maximum amount of context you want to allow for each parallel session. By using a lower maximum, you don't have to allocate as much VRAM for each session, allowing more usage for the same amount of hardware. Thus, cheaper.

sergiotapia 1 day ago

I can't seem to find pricing for this model. Since the context size is just a quarter of the full size K3, is the price also much cheaper?

I usually keep my context in chats below 256k anyways so this would be tremendous honestly.

  • daemonologist 1 day ago

    It seems to only be available in Kimi Code, via subscription, no there's no API pricing. The linked page says it consumes about half as much quota as the 1M version though.

    • trollbridge 17 hours ago

      It's not available via API (yet).

jscott817 23 hours ago

Can we assume that model performance at 90% of the 256k limit != 90% of 1M token limit?

Is this the exact same model just with less VRAM allocated for context window?

  • 3836293648 11 hours ago

    No, you can't assume it. You can trust them as they made that claim outright. Or you can choose to not trust them, I guess

gigatexal 22 hours ago

Not relevant to this link but I was thinking about the allegations of Chinese AI companies distilling from the big frontier American ones. And I came to the conclusion: I don’t care.

Who cares? China has always copied and then copied the means of production and then out produced. See also Tesla and now all the Chinese cars eating their lunch.

As long as I get really solid AI models for cheap that do what I need I don’t care if they’re Chinese or otherwise.

I’ll still never use Grok from SpaceX AI cuz eww no, I have principles. ;-)

illithid0 1 day ago

This was posted 38 minutes ago, and as of 20 minutes ago, several Anthropic services are now designated as having a "major outage".

Doubt these are related, but it made me laugh a little.

  • paxys 1 day ago

    Anthropic services have outages on all days ending in y.

    • de6u99er 1 day ago

      So users in Germany are not affected?

      • zarmin 23 hours ago

        Germany ends in y

        • tshaddox 23 hours ago

          Not in Germany.

          • hebelehubele 23 hours ago

            The word "Germany" is "Germany" everwhere and always ends with y. The country Germany has different names though.

            • aksappy 23 hours ago

              Germans prefer Deutschland, no?

              • 361994752 22 hours ago

                so Deutschland is not affected. Germany is

        • vidarh 21 hours ago

          But Tag doesn't.

      • jampa 23 hours ago

        Anthropic has better SLA in Germany. I’ve heard uptime there can get up to nein nines.

holoduke 1 day ago

A bit of topic. But how likely is it that the US will restrict Chinese open weight models and also force Euro countries to do the same? I think it will be effective within 6 months. The US is having a hard time staying competitive.

  • jeppebemad 1 day ago

    I don’t know, but I do think that the days of the US “forcing” Euro countries to do anything, is over.

    • holoduke 1 day ago

      No it's not. The US dictates every single step in Europe. Euro politicians are in the pockets of American institutions. The biggest reason of dumb decisions in the EU is because of deliberate decisions made in favour of the US. LNG, war in Ukraine, ASML export restrictions and many more.

  • realusername 1 day ago

    Trump burned a lot of bridges in the EU, that one will be a hard sell

  • impossiblefork 23 hours ago

    It isn't legally possible for them to do this at the EU level. The EU parliament would never vote for it.

    For pressure at the country level leading to this kind of thing I think it's very unlikely. Here in Sweden it wouldn't just require a vote in the Swedish parliament and before this there'd have to be förarbeten and you can't just brazenly push things through with insane arguments, Swedish social convention goes against it-- and there's just no way to get it through.

    It also might not even be legal. "We aren't at war with China and I'm a communist, and the US LLMs are so aligned with values inimical to my political ideology that this is interference with opinion formation" might be an actual legal argument that the ECHR or CJEU might actually have to accept.

    • holoduke 23 hours ago

      It happened with ASML. What's your thought on it? The US forbid dutch company to export.

      • impossiblefork 23 hours ago

        Well, that's the deal, I assume-- that they weren't allowed to buy the Japanese light source outright, so they bought the American one, even though it required giving the Americans some sort of veto or control.

        I guess it sucks if one wants to expert broadly, but if you're big on vertical integration and the Japanese won't sell I guess you take what you can get.

dools 23 hours ago

Hopefully this helps reduce some of the pressure on their infrastructure. Their models have all become super dumb recently and their support are not addressing it. I have a hunch they’ve been serving a significant percentage of requests with quantised models.

  • stingraycharles 22 hours ago

    Ah come on, HN is not the place for these types of unfounded conspiracy theories that keep popping up on Reddit.

periodjet 1 day ago

Why are Anthropic and OpenAI even allowing their coding harness apps to be plugged into different model providers…? I’m surprised they haven’t figured out a way to clamp down on that by now.

  • wren6991 1 day ago

    I'd guess because it costs them nothing and it gives you a smoother transition back towards paying for their products.

  • auspiv 1 day ago

    Well when a huge part of potential revenue is all in on Bedrock... you need the harness to be able to talk to Bedrock. And Vertex. And all the other places these models are hosted. And allow for proxy because many businesses do not all direct internet access... all valid business reasons.

  • archildress 1 day ago

    From my view, as soon as they do that, they send people out the door to use Opencode instead - and once many people have a taste of trying every model via Openrouter, it's eye opening as to the possibilities.

    Of course - Anthropic and OpenAI have an advantage in the amount they can subsidize the usage, but I think those days are waning.

  • HDBaseT 22 hours ago

    It honestly should be easier, its a by product of everyone using the same API standard.

    Codex and Claude require editing a .json file, but most other harnesses have direct connections via a /provider or /login command.