Z.ai has started rolling out its GLM-5.2 model to subscribers on its GLM Coding Plan, while also outlining plans to make the system available through API and chatbot products in the coming week. The company said the model is already accessible to users on Lite, Pro, Max, and Team tiers.
In a post shared on social media, the company positioned GLM-5.2 as its new flagship model and framed the release as part of a broader push for open and accessible AI tools. Z.ai said the model is intended to support developers and builders, with an emphasis on coding and long-running tasks.
According to Z.ai, GLM-5.2 offers stronger programming performance and support for up to 1 million tokens of context. The company said that long-horizon work remains one of the model’s strengths, suggesting it is designed for extended problem-solving sessions and tasks that require maintaining continuity over a large amount of information.
Z.ai also said the model includes two reasoning settings, labeled High and Max. For coding use cases, the company recommends using Max, which it says enables deeper reasoning and more reliable output. The post did not include benchmark data or a direct comparison with other models.
The current release is limited to Coding Plan subscribers, but Z.ai said broader access is on the way. API and chatbot services are scheduled to launch next week, which would give developers and general users another way to try the system outside the subscription plan.
The company also said GLM-5.2 will be officially open sourced next week under the MIT License. That would make the model’s weights and code available under permissive terms that allow reuse and modification, subject to the license conditions.
Open sourcing is a notable part of Z.ai’s strategy, which the company has highlighted in past model announcements as well. The new release continues that pattern by combining early access for paid users with a planned public release.
Z.ai’s post also repeated its broader view that AI should be open, accessible, and available to developers everywhere. The company described the upcoming rollout as part of that mission, though it did not provide a detailed roadmap beyond the planned API, chatbot, and open-source launch.
The announcement places GLM-5.2 in a staged release process, beginning with paid Coding Plan users before expanding to more public channels. That approach is common among AI model developers, who often test new systems with a limited audience before opening them more widely.
For developers already using Z.ai’s tools, the immediate update is access to GLM-5.2 in the subscription product. For others, the key changes are the expected service launches next week. Those additions could broaden the model’s reach if Z.ai delivers the API and chatbot availability as promised.
The company has not said how GLM-5.2 will be priced when it reaches the API or whether the chatbot will be free or paid. It also has not released further technical details in the announcement beyond its coding focus, long-context support, and reasoning levels.
Still, the release signals another step in Z.ai’s effort to position GLM as a developer-facing model family with both proprietary access and open-source distribution. With access for Coding Plan users already live and wider availability scheduled soon, GLM-5.2 is set to become the company’s latest test of that strategy.