OpenCode 2.0
OpenCode is an open source AI coding agent available as a terminal-based interface, desktop app, or web app. It supports many LLM providers, plugins, MCP servers, and custom commands, and can also be installed via Docker images.
Key facts
What is OpenCode 2.0?
An open source AI coding agent that runs in the terminal, desktop, or browser and connects to the LLM providers of your choice.
Who is OpenCode 2.0 best for?
Developers, engineering teams, CLI/terminal users, self-hosters and teams using Docker
What does OpenCode 2.0 cost?
Verified starting price is USD 10 / month. View source
What are its main limitations?
- Windows package managers are not supported (standalone binary or Docker instead)
- Official pricing page currently returns a 404 error
- Easy access to top coding models via OpenCode Console requires a separate product
Limitations are based on product materials or editorial synthesis; confirm on the official site.
When was this information verified?
October 13, 2026
Sources
- Official homepageofficial homepage · Oct 13, 2026
- Official pricingofficial pricing · Oct 13, 2026
Key features
- ✓Terminal-based agent interface (TUI) with themes and keybindings
- ✓Desktop app builds for macOS, Windows, and Linux (.deb, .rpm, AppImage)
- ✓Web interface accessed with the `opencode pair` command
- ✓Built-in support for many LLM providers, connected directly in the TUI with /connect
- ✓Custom providers, plugins, commands, MCP servers, and agent configuration
- ✓Multiple install paths: curl script, Homebrew, npm, Bun, pnpm, Yarn, Vite+, and AUR
- ✓Standalone CLI binaries for macOS, Windows, and Linux (glibc and musl, x64/ARM64)
- ✓Docker images with versioned tags (e.g. ghcr.io/anomalyco/opencode:2.0.0)
Use cases
- →Running an AI coding agent directly in the terminal
- →Pairing with the agent through a browser via `opencode pair`
- →Connecting OpenCode to a preferred or custom LLM provider with /connect
- →Extending the agent with plugins, custom commands, and MCP servers
- →Deploying or scripting OpenCode through versioned Docker images
Pricing
OpenCode Go
$10/mo
- · Subscription plan granting access to the best open source models
OpenCode Go is a $10/month subscription granting access to the best open source models; the official pricing page currently returns 404 and no free plan is stated.
Pros
- +Open source AI coding agent
- +Multiple interfaces: CLI, desktop, and web
- +Broad install-method coverage across package managers
- +Connect to many LLM providers, including custom ones
- +Customizable via config, plugins, MCP servers, and commands
Cons
- −Windows package managers are not supported (standalone binary or Docker instead)
- −Official pricing page currently returns a 404 error
- −Easy access to top coding models via OpenCode Console requires a separate product
Positioning
- Core value: An open source AI coding agent that runs in the terminal, desktop, or browser and connects to the LLM providers of your choice.
- Ideal for: Developers, engineering teams, CLI/terminal users, self-hosters and teams using Docker
- Product type: Terminal (CLI) app, desktop app, web app (via `opencode pair`), and Docker images
Frequently asked questions
What is OpenCode?
OpenCode is an open source AI coding agent available as a terminal-based interface, desktop app, or web app.
How much does OpenCode cost?
The documentation mentions OpenCode Go, a $10/month subscription plan granting access to the best open source models. No other pricing is stated, and the official pricing page returns a 404.
Is OpenCode open source?
Yes, the documentation describes OpenCode as an open source AI coding agent.
How do I install OpenCode?
Via the curl install script, Homebrew, npm, Bun, pnpm, Yarn, Vite+, or AUR, or by downloading a standalone CLI binary, desktop build, or Docker image. Windows package managers are not supported.
Which LLM providers can I connect to?
OpenCode has built-in support for many LLM providers that you can connect to directly in the TUI with /connect, and custom providers can be configured. OpenCode Console is also offered for easy access to top coding models.
Is there a web interface?
Yes. The OpenCode web interface is accessed with the `opencode pair` command, which shows a local URL, username, and password.
You may also like
Curated suggestions based on similar categories.
Rescript
ReScript is a strongly typed programming language that compiles to clean, efficient JavaScript. It is designed to be readable by both humans and AI tools, providing a fast compiler and static type system for building JavaScript, Node, and React applications.
LongCat-2.0
LongCat-2.0 is a next-generation, open-source MoE (Mixture of Experts) model with 1.6T total parameters, designed for agentic coding and featuring native 1M token context support. It achieves state-of-the-art performance on software engineering and terminal benchmarks, and is available on longcat.ai and OpenRouter.
VibeCoding
Dharmesh Shah's proposed idea for a community platform called VibeCoding.com, designed to connect vibe coders with professional developers. The free tier allows browsing and learning from others, while the paid tier lets users post issues (e.g., GitHub repos) and receive paid professional help, creating a marketplace.
Cursor
Cursor is an AI-powered coding agent that helps developers build software faster by integrating directly into their workflow. It offers context-aware code completions, autonomous agents, cloud-based execution, and integrations with tools like Slack and GitHub, all while supporting multiple frontier models.
CodeRide
CodeRide is a project management tool for AI coding agents that uses MCP (Model Context Protocol) to break project documentation or PRDs into optimized, fully contextual tasks. It maintains persistent context between coding sessions, eliminating the need to re-explain project architecture, coding patterns, and goals.
Recurse
Recurse is an AI-powered code review tool that integrates with Claude Code and Cursor to automatically catch bugs and suggest fixes. It understands your full codebase and library dependencies, helping developers reduce review time by up to 80%.