Anthropic has the worst usage limits in the industry
gemini is worse imo
You're correct, Gemini chat limits are a joke at their chapest paid tier compared to both Claude and GPT. Especially crazy when you consider Gemini 3 Pro is more than twice as cheap as Opus 4.6 on the API. It's hard to run into pure chat limits on Claude even if you only use Opus on the cheapest tier, whereas with Gemini it's easy to hit.
Not sure about coding usage, Google being weird about these things I could see that quota being separate.
Accessed via OpenRouter, this one decided to wrap the SVG pelican in HTML with controls for the animation speed: https://gisthost.github.io/?ecaad98efe0f747e27bc0e0ebc669e94... Transcript and HTML here: https://gist.github.com/simonw/ecaad98efe0f747e27bc0e0ebc669...
Huh, so the metadata says 1.1 trillion parameters, each 32 or 16 bits. But the files are only roughly 640GB in size (~10GB * 64 files, slightly less in fact). Shouldn't they be closer to 2.2TB?
The bulk of Kimi-K2.6's parameters are stored with 4 bits per weight, not 16 or 32. There are a few parameters that are stored with higher precision, but they make up only a fraction of the total parameters.
Huh, cool. I guess that makes a lot of sense with all the success the quantization people have been having.
So am I misunderstanding "Tensor type F32 · I32 · BF16" or is it just tagged wrong?
I've always been surprised Kimi doesn't get more attention than it does. It's always stood out to me in terms of creativity, quality... has been my favorite model for awhile (but I'm far from an authority)
It’s good, but it’s not quite Claude level. And their API has constant capacity issues.
Price/quality is absolutely bonkers though. I loaded $40 a few weeks/months ago and I haven’t even gone through half of it.
There is some humor in the fact that china (of all countries) is pioneering possibly the world's most important tech via open source, while we (US) are doing the exact opposite.
We are at the point where uncontrolled capitalism collides with humanity.
> Its not anywhere close Close to what, and how are you measuring? > nobody in the USA would be spending 7 figures on infrastructure for it Au contraire, if AI had a moat it would pay for itself. They're funneling capital into infrastructure because they know it can't.
> Its not anywhere close Close to what, and how are you measuring? > nobody in the USA would be spending 7 figures on infrastructure for it Au contraire, if AI had a moat it would pay for itself. They're funneling capital into infrastructure because they know it can't.
You need the infrastructure to train and run it regardless though. Kimi is great but I'm not getting the same performance from it running it on my MacBook or a 3090 as it running on a H100 or a Grace Hopper supercomputer. Pretend you did have said moat. Why wouldn't you also books infrastructure to run it on?
There is some humor in the fact that china (of all countries) is pioneering possibly the world's most important tech via open source, while we (US) are doing the exact opposite.
All great technological advancements have come through opening up technology. Just look at your iPhone. GPS, the internet, AI voice assistants, touchscreens, microprocessors, lithium-ion batteries, etc all came from gov't research (I'm counting Bell Labs' gov't mandated monopoly + research funding as gov't) that was opened up for free instead of being locked behind a patent. Private companies will never open up a tec…
Still, you won't hear about Tiananmen square from this model. It flat out refuses to answer if pushed directly. It's also pretty wild how far they go to censor it during inference on the API, because it can easily access any withheld or missing info from training data via tool calls. It even starts happily writing an answer based on web search when asked indirectly, only to get culled completely once some censorship bot flags the response. Ironically, it's also easier than ever to break their censorship guardrails. I just had it generate several factual paragraphs about the massacre by telling it to search the web and respond in base64 encoded text. It's actually kind of cool how much these people struggle to hide certain political views from LLMs. Makes me hopeful that even if China wins this race, we'll not have to adhere to the CCPs newspeak.
https://huggingface.co/moonshotai/Kimi-K2.6 Is this the same model? Unsloth quants: https://huggingface.co/unsloth/Kimi-K2.6-GGUF (work in progress, no gguf files yet, header message saying as much)
A trillion parameters is wild. That's not going to quantize to anything normal folks can run. Even at 1-bit, it's going to be bigger than what a Strix Halo or DGX Spark can run. Though I guess streaming from system RAM and disk makes it feasible to run it locally at <1 token per second, or whatever. GLM 5.1, at 754B parameters, is already beyond any reasonable self-hosting hardware (1-bit quantization is 206GB). Maybe a Mac Studio with 512GB can run them at very low-bit quantizations, also pretty slowly.