GLM-5.2 pairs a 1M context window with MIT open weights
- Source
- x.com
- Date
Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong long-horizon capabilities with a 1M context window - Two levels of reasoning effort: GLM-5.2 (max) pushes the limits, while GLM-5.2 (high) strikes a strong balance between performance and token efficiency - MIT-licensed open weights - Same API pricing as GLM-5.1 Tech Blog: https://t.co/LAsxUdN0JZ Weights: https://t.co/g0A1C4UWx4 API: https://t.co/Kc3E22cbN7 Coding Plan: https://t.co/Nk8Y98HNhU Chat:
GLM-5.2 leads GLM-5.1 by a wide margin across various domains, including coding, tool usage, reasoning, and general knowledge.
For GLM-5.2, we strengthened 1M-context training for coding agents across large-scale implementation, automated research, performance optimization, and complex debugging. The result is a long-context system that is both broad in scope and reliable in execution.
We also detail the infrastructure innovations powering our 1M context window and agentic RL. Explore the full breakdown at https://t.co/wDDG7PloPE.
Context
Z.ai's GLM-5.2, announced June 16, 2026, combines a million- with downloadable model weights under the MIT license. The context window is the amount of material the model can work with in one request, which gives long coding tasks room for more source code and accumulated work. The model card also lists local deployment options such as vLLM and SGLang, so access is not confined to the company's hosted service.
The launch announcement offers two reasoning-effort levels, high and max. Z.ai's own chart, averaging three coding benchmarks run in Claude Code 2.1.167, shows max using roughly twice the average output of high (about 84,000 versus 44,000 tokens per task) for a score only a few points higher. That is vendor evidence, but it illustrates a tradeoff developers can test themselves: more reasoning costs more output even with the same model.
- token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- open weights — A model whose trained parameters are published for anyone to download and run — unlike API-only models you can access but never possess.
postWe’re sharing how GLM-5.3 helped build and optimize the inference infrastructure serving GLM-5.3-Flash.- postGLM-5 lands in Zhipu's Coding Plan Pro as pricing shifts
articleZ.ai details GLM-5.3's leap in exploit analysis and staged release
postGLM-5.3 API launches for coding and cybersecurity agentic tasks
Checking sign-in…
Loading comments…