PCL Logo
CoursesAcademyWhat's New
Get StartedLogin
Open-source AI coding agent

OpenCode

Code with the models and workflow you choose.

An open-source AI coding agent for your terminal, IDE, or desktop. Choose the models you trust and work across sessions in parallel.

Visit OpenCodeRead the docs
academy-project

$ opencode

Install in your terminal

curl -fsSL https://opencode.ai/install | bash

01explore
02build

Open-source AI coding agent

Keep control of the tools, models, and context behind your work

Bring your model

Connect Claude, GPT, Gemini, local models, and more through a provider-neutral workflow.

Work in parallel

Run multiple focused agent sessions on one project while you compare approaches and review the work.

Code-aware context

Use language servers to give agents the symbols, definitions, and references behind your code.

Share the trail

Share a session link when you need feedback, want to explain a decision, or document a debugging path.

A deliberate workflow

Explore, delegate, then review

OpenCode makes it practical to keep research, implementation, and review in focused sessions without giving up project context.

Read the docs
terminal

$ opencode

Install in your terminal

curl -fsSL https://opencode.ai/install | bash

  1. 01Ask an agent to map the codebase with language-server context.
  2. 02Run a second session to prototype or challenge the implementation plan.
  3. 03Review the diff, tests, and shared session trail before you ship.

Try it on a real task

Build an evidence-based AI-assisted workflow

  • 01

    Compare two approaches

    Ask separate sessions to implement and critique two designs, then review the trade-offs before choosing one.

  • 02

    Map an unfamiliar repository

    Use code-aware questions to trace a feature from route to service before making your first change.

  • 03

    Debug with evidence

    Give an agent a failing test or error, inspect its reasoning, and turn the final session into your debugging notes.

Make your development workflow yours

Start with the terminal, connect the models you use, and build a practice you can inspect at every step.

Visit OpenCode
PCL

Master the future of technology through hands-on GPU-accelerated curriculum.

Platform

All Courses

Support

Terms of ServicePrivacy Policy
(c) 2026 PCL Platform. All rights reserved.Practice-centered AI education for modern classrooms.