GLM-5.3 Flash Released: Zhipu's New AI Model Beats GLM-5.2
Zhipu officially releases GLM-5.3 Flash, scoring 84.3 on Terminal-Bench, outperforming GLM-5.2 at 10x lower cost. Open weights, MIT license, 1M context.
Meeting the new king — GLM-5.3 Flash.
Zhipu officially presented and released GLM-5.3 Flash. It was also confirmed that the anonymous Ox Alpha, which had previously spread across OpenCode and OpenRouter, was exactly this model.
The official release has become more powerful and noticeably more stable.
According to Z.ai, GLM-5.3 Flash scores 84.3 on Terminal-Bench 2.1, almost catching up with Claude Opus 4.8 with a score of 85.0.
At the same time, it outperforms the significantly larger GLM-5.2 in all six published tests and costs about 10 times less to use.
Input — $0.15 per million tokens.
Output — $0.50.
Cached input — $0.03.
Open , MIT license, built-in multimodality, and 1 million token context.
Comments
Weekly digest
The best of vibe coding, AI agents and open source — once a week, no spam.
Related articles
All articlesCline Desktop Launches AI Agent Workspace for macOS & Windows
Cline Desktop offers an open‑source AI workspace with parallel agents, task scheduling, checkpoints, branching, and a plugin marketplace, now on macOS and Windows.
Graph-Based Multimodal RAG for Document Processing on LightRAG
Open-source, graph-based universal multimodal RAG system built on LightRAG to process documents and unify text, images, tables, and more.
Moonshot Releases KIMI K3: The World's Largest Open-Source Model
Moonshot unveils KIMI K3, a 2.8 trillion parameter MoE model with a 1M token context window, outperforming major models in knowledge-based tasks.