Run OpenCode in MXterm: install, first session and a split pane for tests

The short version

1. Install it from the project's own instructions, and check which major version you are installing.

2. Open a new tab in MXterm, choose your project folder and add one pane of it from the agent list.

3. Give the agent a task, then split the pane for a shell.

Install it and check the version

The documentation says: "OpenCode is an open source AI coding agent. It’s available as a terminal-based interface, desktop app, or IDE extension." It lists "API keys for the LLM providers you want to use" among the prerequisites (opencode.ai/docs, read 31 October 2026).

Check which major version you are installing. A notice on that page says: "OpenCode 1 and OpenCode 2 both use the opencode command and cannot be installed side by side under their default command." The page calls the install script "the easiest way to install OpenCode 1" and lists npm, pnpm and Homebrew routes as well.

For Homebrew it says: "Use the OpenCode tap above for V1. Homebrew’s brew install opencode formula currently installs V2." The project's releases page lists both v1 and v2 releases (github.com/anomalyco/opencode/releases, read 31 October 2026), and the README on its dev branch gives V2 install commands. Follow the docs for the version you want, and check the page on the day you install.

Connect a model provider

The same page says: "With OpenCode you can use any LLM provider by configuring their API keys." It tells you to run the /connect command, then go to your project directory and start the agent.

MXterm launches the agents you have installed; it does not install them or connect them to a provider.

Step 1: add a pane from the agent list

Open a new tab in MXterm and choose your project folder. The default agent list leads with Claude Code, Codex, Grok and Gemini CLI, then a plain Terminal row and OMP, and lists every other agent after them, installed ones first. In that default list OpenCode is further down, and the list scrolls.

Each installed agent has plus and minus buttons that set how many panes of it the tab opens; press Create workspace to open the tab. An agent that is not installed shows Not installed and an Install button, which opens the install instructions or shows the command.

Step 2: give it a task

Type the task as you would in any terminal; the pane is a real shell, so the agent runs in it as it would anywhere else.

If no provider is connected yet, run /connect as the docs describe.

Step 3: a shell beside it

Use the split buttons in the pane header (Split Right and Split Down) to put a plain shell next to the agent for your tests. Split panes run real shells.

Run your test command in the shell pane, and keep the agent in the pane beside it.

What MXterm adds

See which agent is working, waiting or done: amber means working, red means blocked and green means done.

MXterm does not include a model, so which models the agent can use is set in the agent, not in MXterm.

Next to other agents

OpenCode can run in a pane beside Claude Code or Codex in the same workspace, and each agent can work on its own branch in its own worktree. That reduces collisions; it does not stop two agents editing the same file.

See Claude Code in MXterm, the guide to running Claude Code and Codex side by side in split panes and isolating agents with git worktrees. To keep agents in one room, see the Control Room. Our list of AI coding agents in 2026 gives the vendors' own descriptions of each one.

When a plain terminal is enough

For a task with this agent, a plain terminal is enough once it is installed and connected to a model provider.

MXterm helps when agents or tasks run in parallel and you want their state and their branches in one window. To try it, use the download page.