Collections

Collections are groups of models picked to help you do a job.

  • RAG Starter Stack Embedding, retrieval, reranking, and dataset records for search, semantic memory, and retrieval-augmented generation.
  • Local LLM Starter Stack Small, local-run-friendly model files, GGUF/ONNX/safetensors records, and local inference tools.
  • Global Model Labs Stack Official model records from major Chinese, Eastern, and Western AI labs with license and mirror-readiness context.
  • Agent Toolchain Stack MCP servers, agent frameworks, tool-use projects, eval harnesses, and workflow automation records.
  • Commercial-Safe AI Stack Permissive-license model, dataset, app, and eval records to review first for product work.
  • Verified Downloads Stack Hugging Bay hosted files with SHA-256 hashes, manifests, and resumable download routes.
  • Research Code Bench Stack Paper, GitHub, benchmark, dataset, and model records that connect research context to buildable AI artifacts.

Catalog rows are temporarily unavailable.