DeepSeek-Reasonix

Project Overview

DeepSeek-Reasonix is a DeepSeek-native AI coding agent for your terminal. Engineered around prefix-cache stability for long-running operation. Four ways in — terminal, desktop app, browser, or your editor over ACP.

Key Features

• Config-driven: providers, the agent, enabled tools, and plugins are all declared in reasonix.toml
• Multi-model & composable: DeepSeek ships as a preset; any OpenAI-compatible endpoint is a config entry. Optionally run two models together (executor + planner)
• Plugin-driven: MCP servers and Extension Protocol v1 sidecars
• Cache-aware context maintenance: startup injects stable environment summary, stale tool output is pruned before compaction
• Zero-friction distribution: CGO_ENABLED=0 single binary; cross-compile to six targets

Target Audience

• Developers who want AI coding assistance in the terminal
• Users who need long-running AI agents
• Technical users who value flexible model configuration

Quick Start

1. Install via npm: npm i -g reasonix
2. Or via Homebrew (macOS): brew install esengine/reasonix/reasonix
3. Download prebuilt archives from GitHub Releases
4. Configure reasonix.toml and launch the TUI: reasonix

Core Advantages

• Engineered for cache stability, suitable for long-running tasks
• Single binary distribution with no extra dependencies
• Supports multiple models and plugin extensions
• MIT licensed, open-source and free

System Requirements

• OS: macOS, Linux, Windows
• Architectures: x86_64, arm64
• Additional Dependencies: None (single binary)

Notes

• API keys need to be configured separately
• Desktop app and VS Code extension are also available