00:00:00 - 00:00:52
Last week an anonymous mystery model named Kibitz showed up randomly on Sha Bot Arena performing at the frontier level. A Kibitz is the Yiddish word for that guy standing behind you at a chess game and giving you tips on how to blunder your queen, which to be honest is the perfect analogy for a coding agent. But it was pretty obvious he was behind this mystery model from the beginning because the first two letters of this word are a dead giveaway. Everybody knew it was Kimmy. Then right on cue last Friday, Moonshot dropped

00:00:25 - 00:01:18
Kimmy K 3.1 officially giving us a new 3 trillion parameter open weight model that will be fully downloadable on November 4th, or so they say. And just when you thought the AI bubble was popping and the trust me bro benchmarks could not possibly get any better, Kimmy made them better. >> I can go lower. >> In today’s video, we’ll find out why Kimmy K 3.1’s performance should be impossible and put it to the test on the only benchmark that matters, building 3D horses with 3.js. It is October 26, 2026

00:00:53 - 00:01:41
and you’re watching the Code Report. Earlier this year, Anthropic accused Moonshot AI, the company behind Kimmy, alongside a bunch of other Chinese companies like Deep Seek and Mini Max of running large-scale distillation attacks against Claude by sending swarms of fake users to steal Claude’s output and reasoning tokens, which could then be used to train these open models that they give to us for free. Now, I believe distillation is a fundamental human right because big AI companies have

00:01:17 - 00:02:06
literally distilled the collective human brain by taking our internet web logs, our GitHubs, our scientific papers, and they even destructively scanned our books without asking. Just Google Project Panama. But Open AI and Anthropic don’t really see distillation as a human right like I do. To them it’s Grand Theft Auto Complete and they’ve invested millions and millions of dollars into stopping it. And just a few weeks ago, Anthropic said it pretty much solved the distillation problem with 90

00:01:41 - 00:02:33
to 99% of distillation attacks from China now being blocked. And over at Open AI, they’ve proved that their tech can even stop small-scale distilleries like PewDiePie. But if that’s true, that means an open Chinese model with Kimmy’s performance should not be possible because for years big AI has insinuated that that the reason Chinese models get so good so fast is that they steal big AI’s outputs. But if we can trust this trust me bro distillation data from Anthropic, that would mean that these

00:02:07 - 00:03:01
trust me bro benchmarks from Kimi should not be possible. Like Kimi K3.1 is within one point of Opus 5.5 on Terminal Bench 4.0. That’s pretty crazy and no other open model is remotely close. And we might even be able to actually trust these numbers because artificial analysis ran it independently and got 67% which is the highest score of any model they tested. But one thing to keep in mind is that the harness plays a big role here. Running different models with different harnesses will produce huge

00:02:34 - 00:03:24
differences in benchmark results. In addition, Kimi also wins on SWE marathon which is a test for multi-hour software requirements. Or in other words, jobs that would take humans weeks of work to finish without AI assistance. That’s all pretty impressive, but there is one big caveat when it comes to coding and specifically cybersecurity. Moonshot deliberately ripped out a bunch of cybersecurity data from its training data set where the goal was to make the model even worse at hacking than pirate

00:03:00 - 00:03:54
software. >> I’m a genius, dude. >> And it worked because internally at Moonshot an unfiltered checkpoint solved 90.5% of their illegal exploit tests. But the final release model only solves 81% and that’s a smart business move on their part because just a few days ago the US banned China’s deep seek model from government use including all federal civilian agencies. And they would just love a good reason to ban all Chinese models across the board to hand Open AI and Anthropic a huge win and

00:03:27 - 00:04:18
make all the insiders even more unbelievably rich than they already are. And that’s why Moonshot really has no choice but to lobotomize its hacking capabilities. But despite that, Kimi still performed on top of all publicly available models on the new cyber index at artificial analysis. Although that’s mostly because Claude and GPT-6 declined every prompt on safety grounds and just took the L. But nowadays, the only AI capability that I really care about is how well it can generate a 3D horse with

00:03:52 - 00:04:54
3js. And to analyze the skill objectively and scientifically, I use a formal benchmark. It’s a state-of-the-art eval I’ve developed after thousands of hours of research, which I call thoroughbench. The way it works is I literally just go into the chat window and type create a realistic interactive 3D horse with 3js and then press enter. From there, I rank them from 0 to 10 based on a proprietary loss function I keep in my brain. Using this technique, Google Argon achieved a 0.69, a GPT-6 Astra achieved a 5.21, and Opus

00:04:22 - 00:05:13
achieved a 6.39. But now it’s time to try Kimi K3.1. I tried to be cheap and use the free tier, but got denied. So now I’m paying 250 yuan per month to show you guys this demo. It was worth it, though, because Kimi produced what is easily the most beautiful horse in this sample. It’s coloring is spot on, it looks realistic, and most importantly, its copulatory organ is a reasonable size. I was blown away by its beauty at first glance, but then I push play on the animation and found a

00:04:48 - 00:05:43
problem. The animations are absolutely terrible. If we look closely, you can see it achieves the animation by simply warping the original mesh, which is a very unsophisticated and half-assed way to animate a 3D object. If we compare that to Claude, the horse itself may not look as beautiful, but it actually built a mesh with 95,000 vertices with each shape tied to one of 35 bones, and all those bones are animated independently. Even Gemini Argon used bones, but for some reason it added an extra one. A

00:05:16 - 00:06:10
skeletal animation is more difficult, but it’s why the horse doesn’t look like it’s being torn apart by a black hole when it walks. So the final score for Kimi K3.1 on thoroughbench is a 4.20, which is pretty dang impressive, but still behind the other frontier models. But if you really want to build an amazing 3D horse or user interface, you need to check out Mobbin, the sponsor of today’s video. I’ve been using Mobbin for over 5 years, and its MCP server now lets your coding agent study its library

00:05:42 - 00:06:36
of 600,000 real UI screens before it designs anything. So instead of asking for a login page and getting the same purple gradient vibe slop, you start by telling Claude, Code, Cursor, or Codex to study how the best apps built that complete user flow. It’ll use Mobbin to gather references and build a visual report of the patterns they all share. Then, it’ll recreate that flow and adapt it to your app. Every result points back to the original screen, so you can check the source. And after you connect

00:06:10 - 00:06:35
Mobbin, everything your agent designs will feel like it has actual taste. But, try Mobbin for free today and get a 20% discount at the link below. This has been the Code Report. Thanks for watching, and I will see you in the next one.