[OpenCode Field Notes Part 2] 10-Minute Setup: Install, first run, and baseline config

1. Install

As noted in Part 1, my priority is keeping a stable routine even when tools change. This post is a 10-minute setup aligned with that goal.

Fastest install command:

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

Alternative options:

npm install -g opencode-ai
brew install anomalyco/tap/opencode

Verify install:

opencode --version

2. First run

Run from your project root:

opencode

On first run, connect a provider and choose a default model.

What “provider” means:

Provider: service that offers model API/auth (e.g., OpenAI, Google, Anthropic)
Model: specific model you select inside a provider
Subscription/Auth: payment plan or credentials (API key, OAuth, etc.)

So provider setup is the step where you register which service/model to use with your own credentials.

One big advantage of OpenCode is that you can start with free provider options. Then, with plugins, you can connect existing paid LLM subscriptions in the same environment. In practice: “start free -> expand to paid” within one workflow.

You do not need perfect configuration on day one. Just build a baseline that reduces repeated explanation in long sessions.

Add a primary provider
Add a backup provider
Set a default model
Add one low-cost support model
If needed, connect existing paid channels via plugins

This reduces interruptions when limits or temporary failures happen.

4. Day-one checklist

Install complete and version verified
At least two providers connected
One small test task completed
Run one 30-60 minute session and note where it slows down

Example prompt: “Summarize this project structure first.”

If this is done, you have more than installation. You now have a repeatable starting point for your own workflow.

5. Next

In the next post, I will share my provider strategy for reducing token/subscription friction.

Sources

OpenCode official docs: https://opencode.ai/docs/