Tinygrad Says Kimi K3 Smarter Than GLM-5.3

The official account of the tiny corp posted that Kimi K3 is still a lot smarter than GLM-5.3. The message observes that reinforcement learning aimed at benchmarks stays acceptable when handled carefully and leaves model quality intact without lowering it. At the same time the post states this approach fails to raise core intelligence. The author closes by expressing interest in larger GLM models and the next Kimi release. No other details or confirmations appear in the post.