GLM 5.2

GLM 5.2 is Z.ai's MIT-licensed flagship, a sparse mixture-of-experts model with 753 billion total parameters and roughly 40 billion active per token, supporting a 1 million token input context and 128K output. Its IndexShare architecture reuses a single indexer across every four sparse attention layers, cutting per-token FLOPs by 2.9x at full context length, while speculative decoding via MTP, IndexShare, and KVShare adds a 20% acceptance-length gain. Built for long-horizon agent workflows, project-level software engineering, and complex multi-step automation.