pebrel
AI-native, GPU-accelerated terminal emulator for Windows with SSH, persistent sessions, split panes, and first-class AI CLI workflows.
What is pebrel?
AI-native, GPU-accelerated terminal emulator for Windows with SSH, persistent sessions, split panes, and first-class AI CLI workflows.
Overview of pebrel
What it is
Pebrel (formerly Nebula) is a GPU-accelerated terminal emulator built with Rust and GPUI. It is designed for Windows, macOS, and Linux, and functions as an SSH workspace and a home for AI CLI sessions. The project is open source under the GPL-3.0 license.
Key features
- GPU-accelerated rendering with a native GPUI interface, supporting light and dark themes, backgrounds, and opacity controls.
- Split terminals with draggable panes, saved workspace layouts, and per-pane working directories.
- SSH and SFTP support including host search, proxy and jump-host options, private-key and keyboard-interactive authentication, and host-key verification.
- Intelligent completion that suggests candidates from Git repositories, project scripts, SSH config, filesystem, and registered WSL distributions without requiring a shell plugin or AI request.
- AI CLI workflow integration with dedicated icons and activity states for Claude Code, Codex, and other recognized CLIs; captured answers open in a reader with Markdown, formulas, and local image previews.
- Local and remote file browsing, Git actions, and a Lua configuration system with validation and live reload (TOML configuration remains supported).
Use cases
Pebrel is suited for developers who need a unified workspace for local shells, remote hosts via SSH, file transfers via SFTP, and AI command-line tools. It can be used for managing multiple terminal sessions in tabs and splits, following AI agent activity, and reading output without leaving the application. The intelligent completion feature aids in discovering commands from the current environment, and the SSH host library helps organize and connect to remote servers.
Getting started
See the repository README for download options and installation instructions. Packages are available for Windows (x64), Linux Preview (x64), and macOS Preview (Apple Silicon and Intel).
License and pricing
This project is licensed under GPL-3.0 and is open source.
Based on the project's public README and repository metadata.