Wave Terminal
An open-source terminal that puts editors, browsers and an AI chat in drag-and-drop blocks next to your shell.
Blocks for terminals, editors, previews and Wave AI, with durable SSH sessions and bring-your-own keys.
Wave Terminal is a desktop terminal for macOS, Linux and Windows that treats terminals, editors, web views and an AI chat as blocks you arrange on screen. It suits developers who work over SSH and want an assistant that can read their scrollback without signing up for a hosted agent service.
What it is
Wave AI reads terminal output and scrollback, reads and edits files with backups and your approval, and is reachable from the shell through the wsh ai command. The wsh CLI also copies files between local and remote SSH hosts, and previews cover markdown, images, video, PDFs and CSVs.
Who it's for
It fits developers who juggle remote servers and want files, notes and a chat beside the shell. It is a weaker fit if you want an agent that runs commands for you, because the README lists command execution as coming soon.
Key features
Draggable blocks
Terminals, editors, web browsers and AI assistants are blocks you arrange and toggle to full screen.
Wave AI
A context-aware assistant reads terminal output and scrollback and edits files after you approve.
wsh command line
Manage the workspace from the shell and copy or sync files between local and remote SSH hosts.
Durable SSH sessions
Remote sessions reconnect automatically after network drops and restarts.
Built-in previews
Preview markdown, images, video, PDFs, CSVs and directories without leaving the terminal.
Bring your own models
Use your own OpenAI, Claude, Gemini or Azure keys, or local models through Ollama and LM Studio.
Up and running in 3 steps
Our honest verdict
What's great
- Runs on macOS 11+, Windows 10 1809+ and Linux with glibc 2.28+, on both arm64 and x64.
- Wave AI can use local models through Ollama or LM Studio, so scrollback need not leave your machine.
- The README says no account is required to use it.
- Remote file access and a built-in editor handle files on SSH hosts without a separate sync tool.
Where it falls short
- Command execution by Wave AI is listed as coming soon, so it assists rather than acts.
- Included AI credits are available only during a free beta, per the README, so keys or local models are the long-term route.
- The README gives no RAM, CPU or disk requirements, so you cannot judge small machines from it.
- Building from source needs the steps in BUILD.md rather than a one-command install.
Pricing & license
Wave Terminal is free under the Apache-2.0 licence. The README mentions no paid or cloud plan and gives no prices, only GitHub Sponsors. Your own AI costs are the API keys you add, or nothing if you use local models.
Open remote AI-training roles
880 open remote AI-training roles, updated November 15, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Does Wave Terminal need an account?
The README says no accounts are required. You can bring your own API keys or use local models, and it mentions included AI credits only during a free beta.
Can Wave AI run commands for me?
Not yet. The README lists command execution as coming soon; today Wave AI reads terminal output and files and edits files with your approval.
Which operating systems does Wave support?
macOS 11 or later, Windows 10 1809 or later and Linux with glibc 2.28 or later, on arm64 and x64 builds.
