Such an interesting times we are in,

We just had amazing releases this past two months

kimi k3, glm5.3 qwen3.8 and now glm5.3

These open models are getting really good

Google: Here is the next iteration of our flash model series, with a discount. please use. thx.

Z.ai: Here is our next iteration, neck and neck with Fable/Sol. weights releasing in two weeks.

Is there a plan somewhere that gives access to Kimi K3 and GLM-5.3? I was thinking of testing both to run security reviews of my code.

I know OpenCode Go has both, but their limits seem kinda low, so I'm not sure how feasible it is to run such a task with them.

These results look pretty good, given the smaller model size and the GLM family's historic robustness. Cheaper than Kimi and more robust than DeepSeek. The question in my mind is if you're going cheap, are you going to stop here or go all the way down to DeepSeek Flash?

One htought I had; if The chinese allow unfettered access to cyber capabilties while th US does it's best to neuter it's model releases, from China's point of view they have the US all tied up in knots dealing with problems they don't give people the tools to solve. China giggles as it watches the US under threat from people using it's models. The US is restricting citizens from owning this particular kind of weapon, while China is handing it out to the wrolds citizens freely. It feels like the US would only come out worse overall?

Such a smart model and didn't warn them how confusing the headline is to anyone who understands what "cyber" means?

We got nukes capable of having existential crises, before GTA 6...

They just ignore in their benchmarks opus 5 for some reason :) also grok 4.6 . I wonder why

what i appreciate most about this post is the level of transparency in how they built and scaled an RL pipeline. my friends at the big labs are so cagey about everything, and Zai is just putting out a great crash course for free.