88Z AI, 17 June — Zhipu AI has launched and open-sourced the GLM-5.2 model, which ranks first among globally available models in Code Arena front-end evaluations. Designed for long-horizon tasks, it supports a 1M lossless context window and performs close to Claude Opus 4.8. The model introduces effort level control and an IndexShare architecture to reduce FLOPs, with native adaptation for domestic computing power such as Huawei Ascend.
At a glance
- GLM-5.2 ranks first among globally available models in Code Arena front-end evaluations and supports a 1M lossless context window.
- Reduces computational load via IndexShare architecture and natively adapts to domestic computing power like Huawei Ascend.

The update
Zhipu AI has officially launched and open-sourced the GLM-5.2 model. In Code Arena evaluations focused on front-end development, this model ranks first among globally available models, demonstrating strong coding capabilities. GLM-5.2 is designed specifically for handling long-horizon tasks, supporting a lossless context window of up to 1M to ensure information integrity in complex scenarios. In authoritative benchmarks such as FrontierSWE, its performance is close to that of Claude Opus 4.8. Technically, GLM-5.2 introduces an effort level control mechanism and optimizes floating-point operations (FLOPs) through the IndexShare architecture, improving operational efficiency. Furthermore, the model natively adapts to domestic computing platforms like Huawei Ascend, providing developers with more efficient localized deployment options and further promoting the development and application of the domestic artificial intelligence ecosystem.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- 智谱Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.