With the release of Kimi K3 today, Chinese models are now within striking distance of the best of the US frontier models. I've seen several benchmarks that show K3 at about Opus 4.8/GPT-5.5 level, some even putting it very close to Fable. We used to think that Chinese models were at least 6 month behind the frontier models, but now it looks like maybe 2 or 3 months, if that. MoonshotAI says they'll release the ~3T weights for K3 on July 27th. Of course, to run a 3T weight model you'll need to spend a good $500K on hardware, so not just anyone will be running it, but some companies could decided that renting hardware to run a model like K3 would be worthwhile. And more open weight Chinese models are due to drop over the next month or so: GLM-5.3 and DeepSeek4.1 pro. What's this do to the IPO prospects for Anthropic and OpenAI? Does this help companies that are just running datacenters for those two because now they'll be able to jump in the race themselves?