Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server, 65 tok/s Qwen 3.5 122B, Llama 3.3 70B, Gemma 4 31B. Private, offline, airgap-ready. Built for NDA / legal / healthcare workflows.
Claude Code Local is a Python-based server that runs an Anthropic-compatible API endpoint entirely on Apple Silicon Macs using the MLX framework, allowing Claude Code to communicate with local models instead of Anthropic's cloud. It supports three primary models: Qwen 3.5 122B, Llama 3.3 70B, and Gemma 4 31B, with DeepSeek V4 Flash added via Antirez's ds4 engine for one-million-token context windows. Qwen 3.5 122B achieves 65 tokens per second on-device and completes a Claude Code benchmark task in 17.6 seconds. Four operating modes are available, including a hands-free voice loop that keeps speech recognition and synthesis local. Because no data leaves the machine, the project targets lawyers, healthcare providers, accountants, and contractors who handle NDA-protected or regulated material and cannot send content to external APIs. The repository also references a pre-configured Mac mini product called the AirGap Box for teams that prefer a ready-to-deploy appliance.
- ✓Open-source license (MIT)
- ✓Actively maintained (<30d)
- ✓Healthy fork ratio
- ✓Clear description
- ✓Topics declared
- ✓Documented (README)
git clone https://github.com/nicedreamzapp/claude-code-localTools overview
What people ask about claude-code-local
What is nicedreamzapp/claude-code-local?
+
nicedreamzapp/claude-code-local is tools for the Claude AI ecosystem. Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server, 65 tok/s Qwen 3.5 122B, Llama 3.3 70B, Gemma 4 31B. Private, offline, airgap-ready. Built for NDA / legal / healthcare workflows. It has 2.8k GitHub stars and was last updated 7d ago.
How do I install claude-code-local?
+
You can install claude-code-local by cloning the repository (https://github.com/nicedreamzapp/claude-code-local) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.
Is nicedreamzapp/claude-code-local safe to use?
+
Our security agent has analyzed nicedreamzapp/claude-code-local and assigned a Trust Score of 100/100 (tier: Verified). See the full breakdown of passed checks and flags on this page.
Who maintains nicedreamzapp/claude-code-local?
+
nicedreamzapp/claude-code-local is maintained by nicedreamzapp. The last recorded GitHub activity is from 7d ago, with 0 open issues.
Are there alternatives to claude-code-local?
+
Yes. On ClaudeWave you can browse similar tools at /categories/tools, sorted by popularity or recent activity.
Deploy claude-code-local to your cloud
Ship this repo to production in minutes. Each platform spins up its own environment with editable env vars.
Maintain this repo? Add a badge to your README
Drop the badge into your GitHub README to show it's tracked on ClaudeWave. Each badge links back to this page and reflects the live Trust Score.
[](https://claudewave.com/repo/nicedreamzapp-claude-code-local)<a href="https://claudewave.com/repo/nicedreamzapp-claude-code-local"><img src="https://claudewave.com/api/badge/nicedreamzapp-claude-code-local" alt="Featured on ClaudeWave: nicedreamzapp/claude-code-local" width="320" height="64" /></a>More Tools
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
An AI SKILL that provide design intelligence for building professional UI/UX multiple platforms
🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman
AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI, and more). Turn any folder of code, SQL schemas, R scripts, shell scripts, docs, papers, images, or videos into a queryable knowledge graph. App code + database schema + infrastructure in one graph.
A light-weight and powerful meta-prompting, context engineering and spec-driven development system for Claude Code by TÂCHES.
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies