Reach the Z.ai reference editorial team

Who you are reaching

This contact page reaches the independent editorial team behind this reference site, not the upstream Zhipu AI lab or the BigModel platform.

To be unambiguous: this site is zai.gr.com, an independent reference resource. The editorial team here has no relationship with Zhipu AI, the BigModel platform, or the ChatGLM project. If you are trying to reach the upstream lab about your account, an API issue, a billing problem, or a business inquiry, this is the wrong destination. The BigModel platform support portal is accessible after login at the upstream platform and is where those questions belong.

If you have found an error on one of our pages, have a factual tip supported by a public source, want to ask about how we handle a particular topic, or have general feedback about the reference site itself, then this is the right place. The team reads all email sent to hello@zai.gr.com and responds to messages that require a response. High-volume periods mean not every message receives an individual reply, but corrections backed by primary sources are reviewed and actioned on every inquiry.

Submitting a factual correction

Corrections backed by a public primary source are the highest-priority category of message the editorial team receives.

Good corrections have three components: the URL of the specific page with the error, the exact passage or claim you believe is incorrect (quoting the specific text is most helpful), and a link to a public primary source that supports the correction. Primary sources include Hugging Face model cards, the official GitHub repositories, platform documentation, arXiv papers, and official announcements. Secondary sources like blog posts and forum discussions are useful context but are not sufficient on their own for a factual correction to a reference page.

The team aims to review corrections within five business days. If a correction is accepted, the page is updated and a note is added to the revision log. If the team disagrees with the correction, a brief explanation of why is sent back by reply. Either way, the submission is not ignored.

The editorial approach to AI capability claims follows the rigorous sourcing standards outlined in academic AI evaluation methodology. For teams who want context on how responsible AI coverage is structured more broadly, the NIST AI Risk Management Framework provides a useful reference for how to think about factual accountability in AI-related publishing.

Submitting a content tip

Tips about new model releases, undocumented platform changes, and significant benchmark results are welcome if they are backed by a public source the team can verify.

The team tracks official channels actively, but significant developments sometimes surface first in community discussions or through developer testing before they appear in official documentation. If you have noticed something worth covering — a new model generation, a platform change that affects the developer workflow, a benchmark result that changes a practical recommendation — email hello@zai.gr.com with a link to the public source and a brief explanation of why it matters.

The team cannot commit to covering everything submitted as a tip, and cannot respond individually to every submission. Tips that come with a clear public source and a specific explanation of the practical impact for developers are more likely to result in a page update than vague requests to "cover" a topic without supporting materials.

Inquiry types and response windows

Four inquiry categories with their appropriate contact route and expected response window.

Editorial team inquiry types, contact routes, and response windows

Inquiry type	Email / route	Response window	Notes
Factual correction	hello@zai.gr.com — subject: "Correction: [page]"	Up to 5 business days	Include page URL, quoted passage, and primary source link
Content tip or new development	hello@zai.gr.com — subject: "Tip: [topic]"	Reviewed; not guaranteed a reply	Include a public source URL and why it matters for developers
General feedback or editorial question	hello@zai.gr.com — subject: "Feedback: [topic]"	Best-effort; 1–2 weeks	Questions about scope, methodology, or coverage approach
BigModel account, billing, or API access	BigModel platform support portal (not this team)	Not handled here	This editorial team cannot access your account or billing records

"The reference pages here gave my team a clear orientation before we committed to the GLM stack. When I sent a correction about a context window figure, it was updated within three days — that kind of editorial responsiveness is rare for an independent site."