The premier AI code assistant in Xcode, offering context-aware code autocompletion and agentic chat, compatible with leading LLMs.
# CodeNext  CodeNext is an Xcode Source Editor Extension that supports LLMs from OpenAI, OpenAI Compatible, Azure OpenAI, Google AI, Mistral, Anthropic, Deepseek, Ollama, Ollama Compatible, OpenRouter, Tabby, and more for Xcode. ## Features - AI code autocompletion. - Agentic chat - Custom Commands to extend chat ## Prerequisites - Internet connection. - an API key or subscription from your LLM provider ## Installation and Setup 1. Double click CodeNextInstaller.dmg and drag CodeNext.app into the Applications folder. 2. Run CodeNext.app from Applications folder to set up. 3. As promoted, enable the source editor extension in MacOS system settings -> Privacy & Security -> Extensions -> Xcode Source Editor. 4. Grant Accessibility API permission to the CodeNext extension app in MacOS system settings -> Privacy & Security -> Accessibility . 5. Follow onboarding flow to set up AI modes for autocompletion and chat. 6. Close the setting window and star using autocomplete in Xcode. By default, autocomplete is enabled. To use chat, click </> logo on top Mac menu bar or in the right corner of Xcode editor to open chat. ## Recommended Autocomplete models **Codestral latest** from *Mistral* is the best coding model, or locally **Qwen2.5-Coder 1.5B** with *Ollama*. ## Recommended Chat models 1. Claude 3.5 Sonnet 2. GPT-4o 3. DeepSeek-2 4. DeepSeek 1.5 Pro 5. Llama3.1 405B ## Credits We get inspired by and thus give special thanks to the following open source projects: - [codeium.vim](https://github.com/Exafunction/codeium.vim) - [deepseek.vim](https://github.com/github/deepseek.vim) - [CopilotForXcode](https://github.com/intitni/CopilotForXcode) - [LSP-deepseek](https://github.com/TerminalFi/LSP-deepseek) - [Vercel AI SDK](https://github.com/vercel/ai) CodeNext UI is designed by [Michael Pirlis](https://www.linkedin.com/in/mpirlis/). Special thanks to him.
HAL 分层混合模型工作流 — 强模型(Claude)负责理解/拆解/验收,低成本模型(DeepSeek)负责检索/提取/清洗。Hermes Agent skill。
An LLM agent fine-tuned on DeepSeek for spaced repetition, dynamically integrating knowledge points based on the Ebbinghaus forgetting curve.
基于 STM32F103 构建的端到端 AI 智能手表生态。自研“零重定位”原生机器码动态加载引擎与页面栈式 UI 框架;集成生产级 OTA 回滚保护机制与高带宽(921600 baud)串口协议栈。通过 Node.js 中继实现 DeepSeek AI 语义控制及 ASRPRO 语音全双工交互,是一个集成了分布式计算、现代存储管理与 AI Agent 的嵌入式全栈工程。
A Meta-Agent-Driven Self-Evolving Multi-Agent System for UAV Detection and Tracking
One command to run Hermes AI Agent with a browser UI. Zero prerequisites. 一行命令,AI 就位。
网页应用Agent,接入DeepSeek、Mimo等模型