Wow, if the benchmarks checkout with the vibes, this could almost be like a Deepseek moment with Chinese AI now being neck and neck with SOTA US lab made models
Exciting benchmarks if true. What kind of hardware do they typically run these benchmarks on? Apologies if my terminology is off, but I assume they're using an unquantized version that wouldn't run on even the beefiest MacBook?
I pray the benchmark figures are true so I can stop paying Anthropic after screwing me over this quarter by dumbing down their models, making usage quotas ridiculously small, and demanding KYC paperwork.
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)
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.
I think one of the motivations is undermining US companies. OpenAI and Anthropic are the two biggest players, and are American. Open weights models reduce the power those two big players have over the industry. If the Chinese companies tried to play by US rules and close-source their products then people would mostly use ChatGPT and Claude. So the Chinese companies don't make a ton of profit either way, but by releasing the models as open weights they can at least keep the US from making as much profit.
A lot of people speculating on the motivations behind Chinese labs open sourcing their models. The reason is simple and clear: It is the only viable commercialization strategy that is available to them. I wrote about this here: https://try.works/writing-1#why-chinese-ai-labs-went-open-an...
If the benchmarks are private, how do we reproduce the results? I looked up the Humanity's Last Exam (https://agi.safe.ai/) this model uses and I can't seem to access it.
Running it through opencode to their API and... it definitely seems like it's "overthinking" -- watching the thought process, it's been going for pages and pages and pages diagnosing and "thinking" things through... without doing anything. Sitting at 50k+ output tokens used now just going in thought circles, complete analysis paralysis.
Might be a configuration or prompt issue. I guess I'll wait and see, but I can't get use out of this now.
I have been testing it in my app all morning, and the results line up with 4.6 Sonnet. This is just a "vibe" feeling with no real testing. I'm glad we have some real competition to the "frontier" models.
Are there any coding plans for this? (aka no token limit, just api call limit). Recently my account failed to be billed for GLM on z.ai and my subscription expired because of this... the pricing for GLM went through the roof in recent months, though...
I've used Kimi K2.5 when I run out of Codex quota. It does small and medium things OK. But if I work on complex things, I'll later have to spend two days cleaning up the mess with Codex. Hopefully 2.6 does better.
Unfortunately the generation of the English audio track is work in progress and takes a few hours, but the subtitles can already be translated from Italian to English.
TLDR: It works well for the use case I tested it against. Will do more testing in the future.
59 comments
[ 2.6 ms ] story [ 68.0 ms ] threadedit: Note that you can run it yourself with sufficient resources (e.g., companies), or access it from other providers too: https://openrouter.ai/moonshotai/kimi-k2.6/providers
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)
The US is pretty clearly in the collapsing empire phase, we are all just pretending like it isn't happening.
Also discovered that using OpenCode instead of the kimi cli, really hurts the model performance (2.5).
Transcript and HTML here: https://gist.github.com/simonw/ecaad98efe0f747e27bc0e0ebc669...
Might be a configuration or prompt issue. I guess I'll wait and see, but I can't get use out of this now.
I tried it once, although it looks amazing on benchmarks, my experience was just okay-ish.
On the other hand, Qwen 3.6 is really good. It’s still not close to Opus, but it’s easily on par with Sonnet.
Unfortunately the generation of the English audio track is work in progress and takes a few hours, but the subtitles can already be translated from Italian to English.
TLDR: It works well for the use case I tested it against. Will do more testing in the future.