design

Zinc: Local LLM Inference

Run large language models locally with Zinc.
513 stars21 forksZigGuide quality 9/10Updated 9/7/2026100% free · open source
What it does

Zinc allows founders to run large language models locally on their own hardware, specifically AMD GPUs and Apple Silicon, for private and efficient AI inference.

When to use it
  • You need to keep your AI data and models private and offline.
  • Your application requires low-latency and high-throughput AI inference.
  • You want to develop and test AI models on your local machine before deploying them to a cloud or production environment.
Ready-to-paste prompt
Run the command `zig-out/bin/zinc -c config.toml -p 'What is the meaning of life?'` to test your local language model with a sample prompt.
Heads up: Zinc requires a compatible Zig compiler version, specifically Zig 0.10.0 or later, to build and run correctly, so ensure you have the correct version installed before attempting to use Zinc.
Saves to your device
Use with Claude
New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

🛠️ Technical setup

Expect 20–40 minutes in a terminal — or let your AI agent drive it.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/zinc && curl -fsSL https://workflowstacks.com/api/skills/zinc/claude-skill -o ~/.claude/skills/zinc/SKILL.md
Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)
claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How Zinc: Local LLM Inference works
Codeflow
Free to inspect

Zinc: Local LLM Inference is a very large Zig project (~121k lines across 198 code files, plus 6 test files). It is a full software project: use it through its install path rather than reading it end to end. Last commit this month, MIT license, has a test suite.

Size
Very large codebase
~121k lines · 198 code files · days to read — use, don't read
Setup
Developer setup
A real software project. Use it via its install path; don't expect to read it all.
Runs on
Node.js
No API keys detected
Zig 65%TypeScript 12%Metal 8%Cuda 5%Astro 3%
Where to start reading
  1. 1
    README.md
    Start here — what it does and how to install it
  2. 2
    AGENTS.md
    The instructions the AI actually follows
  3. 3
    package.json
    Dependencies and the commands it exposes
  4. 4
    src/main.zig
    Inside src/ — the main logic begins here
What's in each folder
src/Core code — the actual logic504 files
site/Frontend / UI330 files
docs/Documentation31 files
loops/Folder67 files
tools/Helper scripts22 files
scripts/Helper scripts19 files
benchmarks/Evaluations & benchmarks11 files
writing/Folder11 files
READMEHas testsDocumentedCI checksMIT licenseUpdated this month
Quick Actions
Details
Creator
zolotukhin
Language
Zig
Category
design
Published
3/25/2026

Are you the creator of this tool? Claim your listing → and earn 85% of every sale.