Hugging Face Local Models

What it does

When your agent uses it

• Picking a GGUF model that fits your laptop or GPU memory
• Launching a local OpenAI-compatible server with llama-server
• Finding the exact GGUF file in a Hugging Face repository
• Converting a Transformers model to GGUF when no quantized files exist

Example prompts

• “Find a coding model on Hugging Face that runs under llama.cpp on my MacBook with Metal.”
• “Start llama-server with a Q4_K_M quant of this repo and check that it answers on port 8080.”
• “List the GGUF files in this Hugging Face repo and tell me which quant suits a CUDA card.”
• “Convert this model, which has no GGUF files, so I can run it locally.”

Requirements

• llama.cpp installed (Homebrew, winget or a source build)
• Hugging Face login for gated repositories
• Network access to the Hugging Face Hub

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

1. 1Search the Hub with apps=llama.cpp.
2. 2Open https://huggingface.co/?local-app=llama.cpp.
3. 3Prefer the exact HF local-app snippet and quant recommendation when it is visible.
4. 4Confirm exact .gguf filenames with https://huggingface.co/api/models//tree/main?recursive=true.
5. 5Launch with llama-cli -hf : or llama-server -hf :.
6. 6Fall back to --hf-repo plus --hf-file when the repo uses custom file naming.
7. 7Convert from Transformers weights only if the repo does not already expose GGUF files.

What it can do on your machine

Read from SKILL.md and the folder at commit ca0325b. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • hf
  • brew
  • winget
  • git
  • make
  • python
  • curl

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Hosts in commands or code, which the agent is likely to contact:

  • huggingface.co
  • github.com

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Hugging Face Local Models loads about 945 tokens when it runs, and up to ~4k if it reads all its reference files. Until then it costs about 62 tokens; SKILL.md has 265 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~62

When it runs · the whole SKILL.md, loaded when a task matches

~945

With references · SKILL.md plus every file in references/, read only if the agent opens them

~4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from huggingface/skills at commit ca0325b, republished under its Apache-2.0 licence (© huggingface). 265 words, ~945 tokens.

name

huggingface-local-models

description

Use to select models to run locally with llama.cpp and GGUF on CPU, Mac Metal, CUDA, or ROCm. Covers finding GGUFs, quant selection, running servers, exact GGUF file lookup, conversion, and OpenAI-compatible local serving.

brew install llama.cpp
winget install llama.cpp

git clone https://github.com/ggml-org/llama.cpp
cd llama.cpp
make

hf auth login

https://huggingface.co/models?apps=llama.cpp&sort=trending
https://huggingface.co/models?search=Qwen3.6&apps=llama.cpp&sort=trending
https://huggingface.co/models?search=<term>&apps=llama.cpp&num_parameters=min:0,max:24B&sort=trending

llama-cli -hf unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_M
llama-server -hf unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_M

llama-server \
    --hf-repo unsloth/Qwen3.6-35B-A3B-GGUF \
    --hf-file Qwen3.6-35B-A3B-UD-Q4_K_M.gguf \
    -c 4096

hf download <repo-without-gguf> --local-dir ./model-src
python convert_hf_to_gguf.py ./model-src \
    --outfile model-f16.gguf \
    --outtype f16
llama-quantize model-f16.gguf model-q4_k_m.gguf Q4_K_M

llama-server -hf unsloth/Qwen3.6-35B-A3B-GGUF:UD-Q4_K_M

curl http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer no-key" \
  -d '{
    "messages": [
      {"role": "user", "content": "Write a limerick about exception handling"}
    ]
  }'

© huggingface, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (references) in skills/huggingface-local-models of huggingface/skills.

• SKILL.md
• references/hardware.md
• references/hub-discovery.md
• references/quantization.md

Open the folder on GitHubat commit ca0325b

Used in 3 other repositories

We found 8 copies of this SKILL.md (exact, near-identical or edited) in other folders, from 3 other GitHub owners. This page covers the copy in huggingface/skills, which our catalogue first saw on October 7, 2026.

• waybarrios/opencode-power-packskills/huggingface-local-models
• sickn33/agentic-awesome-skillsskills/huggingface-local-modelsplugins/agentic-awesome-skills/skills/huggingface-local-models+1 more
• lingxling/awesome-skills-cnhuggingface-skills/skills/huggingface-local-modelsantigravity-awesome-skills/skills/huggingface-local-models+2 more