aistackregistry.com - zsh
┌────────────────────────────────────────────────────────────┐ _ ___ ____ _____ _ ____ _ __ / \ |_ _| / ___|_ _|/ \ / ___| |/ / / _ \ | | \___ \ | | / _ \| | | ' / / ___ \ | | ___) || |/ ___ \ |___| . \ /_/ \_\___| |____/ |_/_/ \_\____|_|\_\ REGISTRY 0.1.2 • dependency compatibility • model specs └────────────────────────────────────────────────────────────┘

start here

AI Stack Registry

Fresh package and model facts. Start with /index.json, follow its URLs, or browse /latest/models/index.md.


Models

13 models across 4 providers: anthropic, google, openai, and xai.

Public docs

Snapshots are dated, versioned, and built from listed package registries, provider docs, pip constraints semantics, and uv resolver docs. Current providers: anthropic, google, openai, and xai. Pin a date or use /latest/.

Providers ship new model IDs, limits, and defaults, so tools should fetch the registry during a run. Fetch /index.json, follow the lookup URLs, and use /latest/models/index.md when a human needs the readable catalog. The repo is private for now. Public contract version: 0.1.2.

FAQ

Why does this exist?
Tools can guess when their built-in data is stale. This registry gives them current data to fetch.
What does AI Stack Registry actually provide?
Dated snapshots, a /latest/ alias, model catalogs, and dependency compatibility files. Tools and people can fetch current facts instead of guessing.
How should a tool use it?
Point it at https://aistackregistry.com/index.json. Have it follow the returned model, provider, and package links. Treat the snapshot date and build time as part of correctness.
How does this avoid stale answers?
Fetch metadata at runtime, cite fetched artifact URLs, and fail if the data is missing.
What's intentionally out of scope (for now)?
This is not a framework tutorial site yet. It is a facts registry.