Z.AI Coding Plan

Z.ai announced GLM-5.3-Flash on August 26, 2026 as the first natively multimodal model in the GLM-5 series, explicitly offering it through the Z.ai Coding Plan as a call endpoint. The model is a 320B-total-parameter Mixture-of-Experts design with only 18B active parameters, and introduces a hybrid sparse + linear atten On benchmarks, GLM-5.3-Flash scores 57 on the Artificial Analysis Intelligence Index v4.1.1 at a discounted $0.045 per task — a level previously only available at roughly 10x the cost, and a level of intelligence the vendor positions as a competitive default for many workloads. On coding and agentic evaluations it cons

Z.ai introduced GLM-5.2 on June 16, 2026 as its flagship model for long-horizon tasks, with the Z.ai Coding Plan listed as an explicit call endpoint alongside the chat, GitHub, and HuggingFace surfaces. The release is positioned as a substantial leap over its predecessor GLM-5.1 and is the first Z.ai model to deliver a The technical architecture changes in GLM-5.2 include a new IndexShare mechanism that reuses the same indexer across every four sparse-attention layers, reducing per-token FLOPs by 2.9x at a 1M context length, and an improved MTP (Multi-Token Prediction) layer for speculative decoding that increases acceptance length b

Z.AI released GLM-5.3 on 2026-08-14 as a post-training-only update over the GLM-5.2 base model, accessible via the Z.ai Coding Plan. The company says scaling post-training on its long-horizon environments (IndexShare, SAO, slime stack) produced a 50% gain on its in-house Z.ai Code Bench and open-source SOTA on Terminal An emergent cyber capability is highlighted, with GLM-5.3 scoring 84.5 on CyberGym for vulnerability discovery and more than doubling GLM-5.2 on exploitation benchmarks (ExploitGym 2h/6h of 105/130 and ExploitBench 54.4 vs. 24.4). Open weights are promised about two weeks after launch pending safety evaluation and hard