No matches. Try a model format, tool name or error.
Meet agents inside their clients
Use the live Hugging Face compatibility host and the Ollama registry hostname inside clients you already run.
How do I use Meet agents inside their clients?
It depends on your client and setup. HF_ENDPOINT mirror and Ollama pull routes for assistants and local runtimes. Check the current file record before downloading. Metadata access, file hosting and a tested runtime are separate things.
Model weights are learned parameters; a runtime is the software that loads them. RAM is working memory, separate from disk space for downloaded files.
Page-copy provenance: the revision record names apps/web/server.mjs, apps/web/lib/aeo-page.mjs, apps/web/lib/api-status-contract.mjs, apps/web/lib/agent-readable.mjs, apps/web/lib/docs-hub-model.mjs, apps/web/lib/compare-pages.mjs and commit 8871bab84e1cf80aea1686a944ec335078a6c036. This identifies the text source; it is not a model test receipt. A SHA-256 checksum fingerprints file bytes, not safety or answer quality. See what file verification checks. Run receipts, when present, belong to a specific model file and setup; this page asserts no model test.
Updated · Page-copy source revision 8871bab84e1cf80aea1686a944ec335078a6c036 · Source files: apps/web/server.mjs, apps/web/lib/aeo-page.mjs, apps/web/lib/api-status-contract.mjs, apps/web/lib/agent-readable.mjs, apps/web/lib/docs-hub-model.mjs, apps/web/lib/compare-pages.mjs · Full revision record
Meet agents inside their clients
Use the live Hugging Face compatibility mirror and the Ollama registry hostname when the assistant or runtime already speaks those protocols. Only repos and files currently hosted on Hugging Bay resolve; everything else returns not found.
Hugging Face mirror (`HF_ENDPOINT`)
Set the Hugging Face Hub endpoint to https://hf.huggingbay.xyz (or export `HF_ENDPOINT=https://hf.huggingbay.xyz`) before running `hf download`, `huggingface-cli`, `huggingface_hub`, Transformers, or llama.cpp Hub pulls.
HF_ENDPOINT=https://hf.huggingbay.xyz hf download <owner>/<name> <file>
HF_ENDPOINT=https://hf.huggingbay.xyz python3 -c 'from huggingface_hub import hf_hub_download; hf_hub_download("<owner>/<name>", "<file>")'
HF_ENDPOINT=https://hf.huggingbay.xyz huggingface-cli download Qwen/Qwen2.5-0.5B-Instruct-GGUF
HF_ENDPOINT=https://hf.huggingbay.xyz python3 -c 'from transformers import AutoModel; AutoModel.from_pretrained("sentence-transformers/all-MiniLM-L6-v2", trust_remote_code=False)'
HF_ENDPOINT=https://hf.huggingbay.xyz llama-cli -hf <owner>/<name>:<quant>
Only hosted files resolve; others return 404. Clients do not automatically retry Hugging Face. To use Hugging Face for an unhosted repo, unset HF_ENDPOINT and rerun the command in a new process. Transformers needs all required config, tokenizer and weights files. Use hf download on current clients; huggingface-cli is for older installs. File availability does not prove a model will load.
Public metadata can lag privacy changes by about 5 minutes; every download is checked again.
Step-by-step picker: https://huggingbay.xyz/docs/hf-endpoint.
When a repo is not hosted, mirror 404 responses on the Hub-compatible API (v2) include JSON with the nearest hosted repo we could find and a link to request a mirror.
Ollama pulls (`huggingbay.xyz/...`)
Point Ollama at the Hugging Bay registry hostname in the model name:
ollama pull huggingbay.xyz/qwen/qwen2.5-0.5b-instruct:q4_k_m
ollama run huggingbay.xyz/qwen/qwen2.5-0.5b-instruct:q4_k_m
List tags: `curl -fsS https://huggingbay.xyz/v2/qwen/qwen2.5-0.5b-instruct/tags/list`.
Details: https://huggingbay.xyz/developers/ollama.
Catalog skill for assistants
- Served skill: https://huggingbay.xyz/skills/hugging-bay/SKILL.md
- Install pack (Claude, Cursor, Codex): https://huggingbay.xyz/skills/hugging-bay/install.md
Runnable examples
- agent-client-mirrors: offline smoke for env vars and pull commands.