Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server. 6 fighters incl. Muse-Glimmer 30B (now multimodal — reads images, abliterated), Gemma 4 31B, Qwen 3.5 122B (65 tok/s), DeepSeek V4 Flash (1M ctx). Private, offline, airgap-ready. Built for NDA / legal / healthcare workflows.
# Add to your Claude Code skills
git clone https://github.com/nicedreamzapp/claude-code-localGuides for using ai agents skills like claude-code-local.
Last scanned: 4/24/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-04-24T06:11:28.460Z",
"semgrepRan": false,
"npmAuditRan": true,
"pipAuditRan": true
}claude-code-local is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by nicedreamzapp. Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server. 6 fighters incl. Muse-Glimmer 30B (now multimodal — reads images, abliterated), Gemma 4 31B, Qwen 3.5 122B (65 tok/s), DeepSeek V4 Flash (1M ctx). Private, offline, airgap-ready. Built for NDA / legal / healthcare workflows. It has 3,257 GitHub stars.
Yes. claude-code-local passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/nicedreamzapp/claude-code-local" and add it to your Claude Code skills directory (see the Installation section above).
claude-code-local is primarily written in Python. It is open-source under nicedreamzapp on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh claude-code-local against similar tools.
No comments yet. Be the first to share your thoughts!
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
If Claude Code just told you "you've reached your usage limit" and gave you a reset time hours away, that's what this is for. You keep working — same Claude Code, same terminal, same project — except the model answering is running on your own Mac.
curl -fsSL https://raw.githubusercontent.com/nicedreamzapp/claude-code-local/main/install.sh | bash
No API key. No second subscription. No waiting until 3pm. It works on a 16 GB MacBook and gets better the more RAM you have — see what runs on your Mac.
Your Mac has a powerful GPU built right into the chip. This project uses that GPU to run massive AI models — the same kind that power ChatGPT and Claude — entirely on your computer, and plugs them into Claude Code so the whole coding experience works offline.
No internet, no subscription, nobody sees your code — and it's the full Claude Code experience: edit files, manage projects, drive your browser, or run a hands-free voice session.
The trick: Claude Code speaks the Anthropic API. Local model servers speak the OpenAI API. So everyone bolts a translation proxy in between — and the proxy is slow and fragile. This server speaks Anthropic natively. One process, zero translations:
| 🐌 What everyone else does | 🚀 What we did |
|---|---|
| Claude Code → Proxy → Ollama → Model | Claude Code → Our Server → Model |
| 3 processes, 2 API translations | 1 process, 0 translations |
| 133 seconds per task | 17.6 seconds per task |
🎯 That one change — eliminating the proxy — made it 7.5× faster.
A real NDA. Llama 3.3 70B. Wi-Fi physically OFF. lsof running live. Watch a 70-billion-parameter model audit a confidential legal document, on-device, with the receipts on screen.
More local-AI demos on the channel:
| Video | What happens |
|---|---|
| 🌌 The Rematch | 4 AI engines build northern lights, 3 fully local — the local challenger painted the best aurora |
| 🏁 Hexagon Shootout | Gemma 31B vs Llama 70B vs cloud Claude, same physics prompt, live counters — 2 of 3 with zero cloud calls |
| 🐳 DeepSeek Three-Way | DeepSeek V4 Flash local beats cloud Claude on wall-clock, same MacBook |
| 🎤 NarrateClaude | Speak to Claude Code, hear replies in a cloned voice — 100% on-device |
| 🏠 Mac mini as home AI | Chat with the Mac mini at home from any browser on any phone |
The question every issue here eventually asks. So we measure it. Agent-12 runs each local model through real agent tasks in a sandboxed working directory, judged by what ends up on the filesystem, never by the model's prose. Temperature 0, fixed caps, fresh sandbox per task, one variable moved per comparison. Tasks, judges, runner and the judge-validation gate are all open at github.com/nicedreamzapp/agent12.
| Model | Easy (12) | time | Hard (8) | time | tok/s |
|---|---|---|---|---|---|
| 🏆 Qwen3.6-35B-A3B (MLX 8-bit) | 12/12 | 64s | 8/8 | 125s | 46 |
| Qwen3-Coder-30B-A3B (MLX 8-bit) | 12/12 | 43s | 7/8 | 392s | 84 |
| Gemma 4 31B (MLX 4-bit) | 11/12 | 92s | 8/8 | 348s | 26 |
DeepSeek V4 Flash (2-bit, 0731 imatrix, ds4) |
12/12 | 203s | 8/8 | 551s | 8.4 |
| Qwen3.8-27B (MLX 8-bit, dense) | 12/12 | 351s | 7/8 | 1156s | — |
| Claude Sonnet 5 (cloud, reference only) | 12/12 | 122s | 8/8 | 131s | — |
All local rows: Apple M5 Max, 128 GB. The headline: the local champion clears the same hard suite as cloud Claude, and does it faster on wall-clock (125s vs 131s) because there is no network in the loop. Read the launch story: I took down six of my own benchmark videos, then built the leaderboard · 71-second video.
Qwen 3.8 reaches 8/8 hard when given an 8000-token budget, but at 16.9× Qwen3.6's wall-clock (full writeup). Muse-Glimmer 30B and Nemotron Omni are listed on the board in a separate vendor-reported section with Meta's and NVIDIA's own published numbers, credited and linked, until they get a real Agent-12 run.
We started with one model. Now we ship a roster — and it's a living lineup: we're builders, this repo is always testing and updating, and new fighters get added the day they drop and benchmarked as we run them. Same MLX server, same Anthropic API — swap one env var and you swap the brain. Plus the ds4 engine for DeepSeek V4 Flash via its own native Metal runtime.
| 🟡 Hermes 4 14B | 🟢 Gemma 4 31B | ✨ Muse-Glimmer 30B | 🟣 Qwen 3.8 27B 🆕 | 🟠 Llama 3.3 70B | 🔵 Qwen 3.5 122B | 🐳 DeepSeek V4 Flash ⭐ | |
|---|---|---|---|---|---|---|---|
| Nickname | The One That Runs On Your Laptop | The Quick One | The Fresh Agent | The Full-Precision Sprinter | The Wise One | The Beast | The 1M-Context Whale |
| Build | 4-bit abliterated | 4-bit IT abliterated | 8-bit abliterated (in-house) | bf16, nothing quantized + DFlash 2 drafter | 8-bit abliterated | 4-bit MoE (A10B) | 2-bit asymmetric (ds4 GGUF) |
| Speed | not benchmarked yet | ~15 tok/s | ~18 tok/s | 36.5 tok/s (9.7 without the drafter) | ~7 tok/s | 65 tok/s 🚀 | ~32 tok/s |
| Params | 14 B dense (Qwen3 base) | 31 B dense | ~30 B | 27 B dense | 71 B dense | 122 B / 10 B active | 284 B / 37 B active |
| Context | 40 K | 128 K | 128 K | 262 K | 128 K | 256 K | 1 M tokens |
| RAM | ~8 GB | ~18 GB | ~30 GB | ~59 GB (55 weights + 4 drafter) | ~70 GB | ~75 GB | ~81 GB |
| Min RAM to run |