Z.ai · 2026-06-13 · major
GLM-5.2 — Z.ai's new flagship coding model with 1M context
Z.ai released GLM-5.2 to every GLM Coding Plan tier today, with a usable 1M-token context, 131K max output, and two thinking-effort levels. The API, chatbot, and MIT-licensed weights are promised for next week.
Z.ai's new coding flagship lands first inside the GLM Coding Plan, with API, chatbot, and open weights set for next week.
Key specs
| Context window | 1M tokens |
|---|---|
| Max output | 131,072 tokens |
| Thinking levels | high / max |
What is it?
GLM-5.2 is the latest model from Chinese frontier lab Z.ai (formerly Zhipu AI). It is the successor to GLM-5.1 and is aimed at long-horizon coding and agent tasks. The release is rolling out to GLM Coding Plan subscribers across the Lite, Pro, Max, and Team tiers.
How does it work?
The model runs with a 1,000,000-token context window and a maximum of 131,072 output tokens. It exposes two thinking-effort levels, high (default) and max, which developers select with the /effort command inside their coding agent. Z.ai integrates GLM-5.2 with Claude Code, Cline, OpenCode, Roo Code, OpenClaw, Kilo Code, Crush, and Goose by overriding the Anthropic model environment variables.
Why does it matter?
GLM-5.1 was already a strong open-weights coding model. GLM-5.2 keeps that recipe and pushes the context window to 1M tokens, useful for whole-repo refactors and multi-hour agent runs. It also lands the same day the US government's order to suspend Claude Fable 5 and Mythos 5 hit, giving developers an alternative open-route option once the MIT weights land next week.
Who is it for?
developers running agentic coding workflows who want a 1M-context, open-route option
Try it
subscribers can switch via settings.json: ANTHROPIC_DEFAULT_OPUS_MODEL: GLM-5.2