GLM-5.2: Built for Long-Horizon Tasks
Get smart on it
GLM-5.2 is a flagship AI model designed to handle long-horizon tasks, which are complex projects requiring sustained reasoning over many steps. The model can process 1 million tokens of context, meaning it can work with much longer documents and code than previous versions, while maintaining quality and reliability in real-world engineering scenarios. This matters because supporting long-context work requires the model to remain effective across extended, complex tasks like large-scale coding projects and debugging, not just accept more tokens theoretically. The model introduces architectural improvements including IndexShare, which reduces computational cost, and effort level controls that let users balance performance against speed and computational expense.
Related news
Vibe coding platform Base44 launches own model as AI startups seek defensibility
Wix-owned vibe coding platform Base44 has started rolling out its own AI model — with hopes that it will eventually outperform frontier models.
DiScoFormer: One transformer for density and score, across distributions
DiScoFormer is a transformer model that estimates both the density and score of a data distribution in a single forward pass without requiring retraining for new distributions. The density describes where data points cluster, while the score, the gradient of log-density, points toward more probable regions and is used in diffusion-based generative models, Bayesian sampling, and scientific simulations. Existing methods force a trade-off: classical kernel density estimation works on any distribution but loses accuracy in high dimensions, while neural score-matching models stay accurate in high dimensions but must be retrained for each new distribution. DiScoFormer significantly outperforms kernel density estimation, cutting score error by 6.5 times and density error by more than 37 times in 100 dimensions, while generalizing to distributions with shapes and complexities not seen during training.
