ABOUT THIS PUBLICATION
Sources first. Clear limits.
Practical decision guides for choosing AI models around real tasks, deployment constraints, and evidence.
Who publishes this site
Model Stack Guide is part of the Trust Mesh publishing network. It is maintained by the same operator as Model Factbook, Model Stack Guide, and Model Workflow Lab. These publications provide different views of a shared subject; links between them are related reading, not independent corroboration.
What we publish
We publish task-based selection frameworks and explain the conditions behind a recommendation.
We distinguish information reported by a model developer, our editorial interpretation, and results from experiments that we actually ran. A protocol is a proposed method, not a completed benchmark. Missing evidence stays explicitly unknown.
Sources and updates
We prefer official repositories, model cards, papers, release notes, and product documentation. Product claims carry a source and a version or date where relevant. Publication dates describe publication, and meaningful revisions receive an updated date. We do not present a live ranking or real-time inventory.
Commercial relationships
The publishing operator also provides GEO and content services. Any article involving a paying client, sponsored placement, or an affiliate relationship must carry a visible article-level disclosure. Commercial relationships do not establish a model's capabilities or performance. These launch articles contain no sponsored model recommendation or affiliate links.
Corrections
To report a factual error, open an issue in the public source repository, identifying the page, disputed statement, and a primary source. Please do not include private information.