crynta/terax-ai

Lightweight (7MB) terminal-first AI-native development environment

What it solves

Terax provides a lightweight, all-in-one development environment that integrates a high-performance terminal, code editor, and agentic AI capabilities into a single workspace. It eliminates the need to switch between separate terminal emulators, IDEs, and AI chat interfaces by unifying them into a native application that supports both cloud-based and local AI models.

How it works

Built using Tauri 2, Rust, and React 19, Terax combines several core components:

  • AI-Native Workspace: An agentic AI side-panel that can read/write files, execute bash commands (with approval), and maintain project memory via a TERAX.md file.
  • High-Performance Terminal: A WebGL-rendered terminal with a native PTY backend supporting multiple shells (zsh, bash, pwsh, etc.) and split panels.
  • Integrated Tooling: A CodeMirror 6 editor with AI autocomplete and diff-based edits, a file explorer with fuzzy search, a git graph for source control, and a built-in web preview pane for local dev servers.
  • Flexible AI Backend: Users can provide their own API keys (BYOK) for providers like OpenAI and Anthropic or connect to local inference engines such as Ollama, LM Studio, and MLX.

Who it’s for

Developers who want a fast, terminal-centric workspace that is deeply integrated with AI agents without sacrificing the performance of a native application or the privacy of local-first AI.

Highlights

  • Agentic AI Workflow: Supports multi-step planning, sub-agents, and the ability to spawn and orchestrate other coding agents like Claude Code.
  • Local-First AI: Full support for offline models via local endpoints, with API keys stored securely in the OS keychain.
  • GPU-Accelerated Terminal: Uses WebGL (moving toward WebGPU) for a smooth, block-based terminal experience.
  • Comprehensive Dev Suite: Includes a built-in git graph, file explorer, and web preview, all within a tiny disk footprint (~7-8 MB).
  • Privacy-Focused: No telemetry, no account requirements, and no data sent to the cloud unless configured by the user.

Related

  • Project
    TurixAI/TuriX-CUA

    TuriX is an open-source computer-use agent that allows AI models to perform real actions on a desktop GUI across macOS, Windows, and Linux without needing app-specific APIs.

  • Project
    OleksandrChekhovskyi/hax

    hax is a lightweight, terminal‑native coding assistant written in C. It talks to many LLM providers (including local llama.cpp and Ollama) via a simple REPL or one‑shot mode, keeping a tiny binary size, low memory use, and full transparency of the request/response transcript.

  • Project
    thinkany-ai/termany

    Termany is a specialized terminal emulator designed to manage multiple concurrent AI coding agent sessions, providing a unified interface for monitoring, reviewing agent output, and handling project workflows.