Merrymake vs. GLM-5.3

Merrymake, located in Aarhus, Denmark, serves as a European think tank and research laboratory focused on artificial intelligence. Our goal is to promote a secure and independent shift towards Artificial Superintelligence (ASI). The primary aim of our organization is to make sure that advancements in superhuman intelligence positively impact all forms of life, humanity as a whole, and each individual person. Our team operates at the cutting edge of technology, treating both the development of AI capabilities and the alignment of those capabilities as interconnected engineering challenges that can be addressed. Our research encompasses the entire spectrum of advanced AI development. We are committed to crafting a future that is truly worthwhile, where technological progress enhances human abilities while consistently upholding values such as well-being, dignity, and individual autonomy. By envisioning this tomorrow, we strive to create a world that empowers everyone to thrive.

GLM-5.3 is Z.ai’s frontier coding model built to improve complex software engineering, long-horizon agent work, and advanced technical reasoning through scaled post-training. The model uses the same base model as GLM-5.2, with performance gains coming from additional post-training environments, more diverse tasks, and expanded compute on the existing training stack. Z.ai’s stack includes IndexShare for efficient long-context processing, SAO for reinforcement learning on long-horizon tasks, and slime for large-scale asynchronous post-training. GLM-5.3 is designed to perform better on work that resembles real engineering tasks rather than short coding exercises. Its training environments include production-style workflows where the model must diagnose bottlenecks, inspect documentation, use codebases, run experiments, implement changes, and produce measurable improvements. The model improves coding performance across public and private benchmarks, including Terminal Bench 3.0, DeepSWE, Agents’ Last Exam, and Z.ai Code Bench. GLM-5.3 also improves token efficiency, producing stronger agentic coding results than GLM-5.2 while using fewer output tokens in Z.ai’s internal evaluations. The model supports three reasoning effort levels, low, high, and max, and no longer supports disabling thinking. Z.ai recommends max reasoning effort for coding tasks, while applications using disabled thinking must migrate to enabled thinking before switching to GLM-5.3. The release also reports emergent cyber capabilities, including stronger vulnerability discovery and exploitation-chain reasoning, with open-weight release planned after safety evaluation and hardening. By combining scaled post-training, long-context infrastructure, long-horizon reinforcement learning, coding-agent workflows, benchmark improvements, reasoning controls, and ZCode integration, GLM-5.3 helps developers and researchers work on demanding coding and agentic tasks.