Skip to content

Select language

Current language: English

OpenCode vs HagiCode: Features and Comparison

OpenCode is an open-source, MIT-licensed AI coding agent for the terminal, desktop and IDE, with 75+ model providers. HagiCode is a workbench that runs OpenCode as an executor and adds proposals, commit tooling and multi-repo context around it.

Short answer

OpenCode is an open-source AI coding agent that runs in your terminal, as a desktop app (beta) or in an IDE. It connects to 75+ providers, including your GitHub Copilot or ChatGPT Plus/Pro login, and can run several agents in parallel on one project. HagiCode is a workbench that starts OpenCode as an executor and adds spec proposals, AI commits, multi-repo context and history on top. It does not replace OpenCode and it does not sell models.

OpenCode already covers a lot on its own, including multiple sessions and a desktop app, so many solo developers should stay with it alone. Add HagiCode if you want each change to go through a written proposal, or to manage several repositories and several agents from one place.

OpenCode vs HagiCode at a glance

AspectOpenCodeHagiCode
What it isOpen-source AI coding agent with build and plan agents, LSP support and share linksAI programming workbench that runs agent CLIs, including OpenCode, as executors
Interface and platformsTerminal UI, desktop app (beta), IDE extension; macOS, Linux, WindowsDesktop app (Windows, macOS, Linux), Docker Compose container, Microsoft Store edition
Pricing and licensingMIT-licensed and free; optional OpenCode Zen (pay-as-you-go) and Go ($10/month, as of October 2026) modelsDesktop and Container editions free; Store edition free base, optional one-time TurboEngine purchase
Model and provider access75+ providers via Models.dev, local models, Copilot and ChatGPT loginsHosts no models; uses the providers configured in OpenCode, or OmniRoute if you configure it
Parallel workMultiple agents can run in parallel on the same projectSeveral sessions side by side, managed as named Agents
Spec and planning workflowPlan agent is read-only and suggests changes; Tab switches to buildOpenSpec proposals: propose, validate, apply, archive
Git and commit handlingLeft to you or to the agent's shell; sessions can be shared by linkAI-written commit messages with repository locking; Git management in the UI
Asking you questionsPermission prompts inside its own interfaceInteractive questions are not available for OpenCode in HagiCode
Setup effortOne install command, then /connect a provider and /initInstall OpenCode first, then HagiCode, then pick OpenCode as the Agent CLI
Best forDevelopers who want an open, provider-flexible agent in terminal, desktop or IDERunning several OpenCode tasks across repos with specs, review and history in one place

What OpenCode does well

Provider freedom, including accounts you already pay for

OpenCode connects to 75+ providers through Models.dev and supports local models through OpenAI-compatible endpoints such as Ollama, LM Studio and llama.cpp.

You can sign in with GitHub to use a Copilot account, or with OpenAI to use ChatGPT Plus or Pro, instead of buying separate API credit.

Build and plan agents

The default build agent has full access for development work. The plan agent is read-only: it denies file edits by default and asks before running shell commands. Press Tab to switch.

A general subagent handles complex searches and multistep tasks and can be invoked with @general.

Several interfaces over one engine

Use the terminal UI, the desktop app (currently beta) or an IDE extension. OpenCode also runs as a server, and opencode acp starts it as an ACP subprocess for editors.

It loads the right language servers automatically, and /init writes an AGENTS.md so the agent learns your project conventions.

Optional hosted models: Zen and Go

Zen is a curated set of models OpenCode has tested for coding agents, billed pay-per-request. Go is a low-cost subscription: $10 per month, with Go Plus at $40 per month, as of October 2026.

Both are optional. The agent works the same with your own keys, and the official docs state it does not store your code or context data.

When to choose OpenCode, and when to choose HagiCode

Choose OpenCode when

  • Choose OpenCode alone if you want an open-source agent and the freedom to switch among many providers or local models.
  • Choose OpenCode alone if your Copilot or ChatGPT subscription is how you already pay for models.
  • Choose OpenCode alone if you are comfortable with its built-in multi-session and desktop app for your daily work.
  • Choose OpenCode alone if you want a plan-then-build habit without a separate proposal tool.

Choose HagiCode when

  • Choose HagiCode if you want each change recorded as a proposal and archived when it ships.
  • Choose HagiCode if an OpenCode task reaches into more than one repository and you want them declared up front.
  • Choose HagiCode if you want AI-written commits and a browser editor beside your sessions.
  • Choose HagiCode if you use more than one agent and want one interface for OpenCode, Claude Code, Codex and others.
  • Choose HagiCode if you want shared reference material (a knowledge base) injected into new tasks.

Using OpenCode together with HagiCode

HagiCode does not bundle or replace OpenCode. It finds your opencode executable, starts a local server for the project (opencode serve on 127.0.0.1), and sends sessions and prompts to it, so your install, providers and config stay as they are.

The model list comes from the providers you have configured in OpenCode, and HagiCode shows a runtime health state while the server starts.

  1. Install OpenCode (for example curl -fsSL https://opencode.ai/install | bash, or npm install -g opencode-ai) and run /connect to add a provider.
  2. Install HagiCode Desktop for Windows, macOS or Linux.
  3. In HagiCode, choose OpenCode as the Agent CLI and pick one of your configured models.
  4. Open a repository and start a session or a proposal.
  5. Expect output and tool calls to stream into the conversation, in the shared tool-call view.
  6. To go back, run opencode in your terminal as before. Your OpenCode config and AGENTS.md are not changed by HagiCode.

What HagiCode adds on top of any agent

Parallel sessions and Agents management

OpenCode can run agents in parallel inside its own interface. In HagiCode each OpenCode run is a named Agent with visible status next to your other agents, so mixed teams are managed in one place.

OpenSpec proposals

OpenCode's plan agent suggests changes; HagiCode's propose, validate, apply, archive flow keeps the plan as a record linked to the diff. Free editions allow 3 concurrent proposals.

AI commit and Code Server

HagiCode writes the commit message from your diff and locks the repository while it commits. The embedded Code Server opens files OpenCode changed in a browser editor.

Monospecs and Vault

Monospecs lists your repositories in one file so OpenCode starts a proposal knowing the whole change surface. Vault keeps reference projects and notes that complement the AGENTS.md OpenCode already reads.

OmniRoute, tool-call view and TokenMonitor

OpenCode already supports many providers; OmniRoute can add one more routing layer shared across your agents. OpenCode's tool calls use the same format as other agents, and an optional TokenMonitor status-bar item shows today's tokens and cost.

The bottom line

  • OpenCode is a free, MIT-licensed, provider-flexible agent with its own desktop app; HagiCode is a workbench that runs it and does not replace it.
  • Stay with OpenCode alone if its terminal, desktop or IDE experience already covers your daily work.
  • Add HagiCode for proposals, commit assistance, multi-repo context and one view across several agents.
  • Your OpenCode config is untouched, so trying HagiCode is a low-risk experiment.

Frequently asked questions

Is OpenCode free?
The OpenCode agent is MIT-licensed and free. You pay for models through your own provider keys, an existing Copilot or ChatGPT login, or the optional Zen (pay-as-you-go) and Go ($10/month as of October 2026) plans. See opencode.ai for current pricing.
How do I install OpenCode?
Run curl -fsSL https://opencode.ai/install | bash, or npm install -g opencode-ai, or brew install anomalyco/tap/opencode. The docs note these methods install OpenCode 1, so check opencode.ai/docs for the current version.
Can I use OpenCode with HagiCode?
Yes. HagiCode lists OpenCode as an Agent CLI. Install it and connect a provider, then select it in HagiCode; HagiCode starts a local OpenCode server for you.
Does HagiCode replace OpenCode?
No. HagiCode is the workbench around OpenCode, adding proposals, commits, multi-repo context and history. It needs OpenCode installed and a provider connected to run.
Does OpenCode work with a Copilot or ChatGPT subscription?
Yes. OpenCode can sign in with GitHub to use a Copilot account, or with OpenAI to use ChatGPT Plus or Pro. Some Copilot models need a Pro+ subscription.
OpenCode vs HagiCode: which should I pick?
Pick OpenCode alone if its terminal, desktop or IDE experience covers your work. Add HagiCode if you want proposals, multi-repo context and one workbench for several agents.

HagiCode

HagiCode is an agentic coding workspace. Rich, convenient UI interactions make planning, agent execution, and project work approachable and easy to navigate, creating an engaging development experience.

Explore an engaging development experience through rich, convenient UI interactions.

Visit HagiCodeRead the documentation
  • SmartOpenSpec workflows turn intent into an executable path, from proposal to design, tasks, and a shipped change.
  • EfficientMulti-agent execution keeps research, implementation, and review moving in parallel, across several repositories and the Agent CLIs you already use.
  • InterestingRich, convenient UI interactions make planning, agent execution, and project work approachable and easy to navigate.

Download for Windows

Get it from Microsoft StoreAll downloads

Turn ideas into shipped software.

HagiCode is an agentic coding workspace. OpenSpec workflows plan the change, multi-agent execution carries it out, and Hero Scenario makes the long run visible, so ideas become shipped software.

Smart
OpenSpec workflows turn intent into a path from proposal to shipped change.
Efficient
Multi-agent execution runs research, build, and review in parallel.
Fun
Rich UI interactions make planning, agent work, and projects easier to follow.